mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-11-22 15:42:02 +00:00
Merge branch 'master' into jepen-multiple-tests
This commit is contained in:
Antonio Andelic
commit
c5fe2c9be7
2
.github/workflows/backport_branches.yml
vendored
2
.github/workflows/backport_branches.yml
vendored
@ -466,6 +466,7 @@ jobs:
|
||||
- BuilderDebTsan
|
||||
- BuilderDebDebug
|
||||
runs-on: [self-hosted, style-checker]
|
||||
if: ${{ success() || failure() }}
|
||||
steps:
|
||||
- name: Set envs
|
||||
run: |
|
||||
@ -504,6 +505,7 @@ jobs:
|
||||
- BuilderBinDarwin
|
||||
- BuilderBinDarwinAarch64
|
||||
runs-on: [self-hosted, style-checker]
|
||||
if: ${{ success() || failure() }}
|
||||
steps:
|
||||
- name: Set envs
|
||||
run: |
|
||||
|
||||
2
.github/workflows/master.yml
vendored
2
.github/workflows/master.yml
vendored
@ -974,6 +974,7 @@ jobs:
|
||||
- BuilderDebTsan
|
||||
- BuilderDebUBsan
|
||||
runs-on: [self-hosted, style-checker]
|
||||
if: ${{ success() || failure() }}
|
||||
steps:
|
||||
- name: Set envs
|
||||
run: |
|
||||
@ -1021,6 +1022,7 @@ jobs:
|
||||
- BuilderBinClangTidy
|
||||
- BuilderDebShared
|
||||
runs-on: [self-hosted, style-checker]
|
||||
if: ${{ success() || failure() }}
|
||||
steps:
|
||||
- name: Set envs
|
||||
run: |
|
||||
|
||||
2
.github/workflows/pull_request.yml
vendored
2
.github/workflows/pull_request.yml
vendored
@ -112,7 +112,7 @@ jobs:
|
||||
StyleCheck:
|
||||
needs: DockerHubPush
|
||||
runs-on: [self-hosted, style-checker]
|
||||
if: ${{ success() || failure() }}
|
||||
if: ${{ success() || failure() || always() }}
|
||||
steps:
|
||||
- name: Set envs
|
||||
run: |
|
||||
|
||||
2
.github/workflows/release_branches.yml
vendored
2
.github/workflows/release_branches.yml
vendored
@ -541,6 +541,7 @@ jobs:
|
||||
- BuilderDebMsan
|
||||
- BuilderDebDebug
|
||||
runs-on: [self-hosted, style-checker]
|
||||
if: ${{ success() || failure() }}
|
||||
steps:
|
||||
- name: Set envs
|
||||
run: |
|
||||
@ -580,6 +581,7 @@ jobs:
|
||||
- BuilderBinDarwin
|
||||
- BuilderBinDarwinAarch64
|
||||
runs-on: [self-hosted, style-checker]
|
||||
if: ${{ success() || failure() }}
|
||||
steps:
|
||||
- name: Set envs
|
||||
run: |
|
||||
|
||||
@ -31,6 +31,8 @@
|
||||
* Add OpenTelemetry support to ON CLUSTER DDL (require `distributed_ddl_entry_format_version` to be set to 4). [#41484](https://github.com/ClickHouse/ClickHouse/pull/41484) ([Frank Chen](https://github.com/FrankChen021)).
|
||||
* Added system table `asynchronous_insert_log`. It contains information about asynchronous inserts (including results of queries in fire-and-forget mode (with `wait_for_async_insert=0`)) for better introspection. [#42040](https://github.com/ClickHouse/ClickHouse/pull/42040) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Add support for methods `lz4`, `bz2`, `snappy` in HTTP's `Accept-Encoding` which is a non-standard extension to HTTP protocol. [#42071](https://github.com/ClickHouse/ClickHouse/pull/42071) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* Adds Morton Coding (ZCurve) encode/decode functions. [#41753](https://github.com/ClickHouse/ClickHouse/pull/41753) ([Constantine Peresypkin](https://github.com/pkit)).
|
||||
* Add support for `SET setting_name = DEFAULT`. [#42187](https://github.com/ClickHouse/ClickHouse/pull/42187) ([Filatenkov Artur](https://github.com/FArthur-cmd)).
|
||||
|
||||
#### Experimental Feature
|
||||
* Added new infrastructure for query analysis and planning under the `allow_experimental_analyzer` setting. [#31796](https://github.com/ClickHouse/ClickHouse/pull/31796) ([Maksim Kita](https://github.com/kitaisreal)).
|
||||
@ -66,8 +68,7 @@
|
||||
* Allow readable size values (like `1TB`) in cache config. [#41688](https://github.com/ClickHouse/ClickHouse/pull/41688) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* ClickHouse could cache stale DNS entries for some period of time (15 seconds by default) until the cache won't be updated asynchronously. During these periods ClickHouse can nevertheless try to establish a connection and produce errors. This behavior is fixed. [#41707](https://github.com/ClickHouse/ClickHouse/pull/41707) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Add interactive history search with fzf-like utility (fzf/sk) for `clickhouse-client`/`clickhouse-local` (note you can use `FZF_DEFAULT_OPTS`/`SKIM_DEFAULT_OPTIONS` to additionally configure the behavior). [#41730](https://github.com/ClickHouse/ClickHouse/pull/41730) ([Azat Khuzhin](https://github.com/azat)).
|
||||
*
|
||||
Only allow clients connecting to a secure server with an invalid certificate only to proceed with the '--accept-certificate' flag. [#41743](https://github.com/ClickHouse/ClickHouse/pull/41743) ([Yakov Olkhovskiy](https://github.com/yakov-olkhovskiy)).
|
||||
* Only allow clients connecting to a secure server with an invalid certificate only to proceed with the '--accept-certificate' flag. [#41743](https://github.com/ClickHouse/ClickHouse/pull/41743) ([Yakov Olkhovskiy](https://github.com/yakov-olkhovskiy)).
|
||||
* Add function `tryBase58Decode`, similar to the existing function `tryBase64Decode`. [#41824](https://github.com/ClickHouse/ClickHouse/pull/41824) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Improve feedback when replacing partition with different primary key. Fixes [#34798](https://github.com/ClickHouse/ClickHouse/issues/34798). [#41838](https://github.com/ClickHouse/ClickHouse/pull/41838) ([Salvatore](https://github.com/tbsal)).
|
||||
* Fix parallel parsing: segmentator now checks `max_block_size`. This fixed memory overallocation in case of parallel parsing and small LIMIT. [#41852](https://github.com/ClickHouse/ClickHouse/pull/41852) ([Vitaly Baranov](https://github.com/vitlibar)).
|
||||
@ -86,6 +87,8 @@ Only allow clients connecting to a secure server with an invalid certificate onl
|
||||
* Fix rarely invalid cast of aggregate state types with complex types such as Decimal. This fixes [#42408](https://github.com/ClickHouse/ClickHouse/issues/42408). [#42417](https://github.com/ClickHouse/ClickHouse/pull/42417) ([Amos Bird](https://github.com/amosbird)).
|
||||
* Allow to use `Date32` arguments for `dateName` function. [#42554](https://github.com/ClickHouse/ClickHouse/pull/42554) ([Roman Vasin](https://github.com/rvasin)).
|
||||
* Now filters with NULL literals will be used during index analysis. [#34063](https://github.com/ClickHouse/ClickHouse/issues/34063). [#41842](https://github.com/ClickHouse/ClickHouse/pull/41842) ([Amos Bird](https://github.com/amosbird)).
|
||||
* Merge parts if every part in the range is older than a certain threshold. The threshold can be set by using `min_age_to_force_merge_seconds`. This closes [#35836](https://github.com/ClickHouse/ClickHouse/issues/35836). [#42423](https://github.com/ClickHouse/ClickHouse/pull/42423) ([Antonio Andelic](https://github.com/antonio2368)). This is continuation of [#39550i](https://github.com/ClickHouse/ClickHouse/pull/39550) by [@fastio](https://github.com/fastio) who implemented most of the logic.
|
||||
* Improve the time to recover lost keeper connections. [#42541](https://github.com/ClickHouse/ClickHouse/pull/42541) ([Raúl Marín](https://github.com/Algunenano)).
|
||||
|
||||
#### Build/Testing/Packaging Improvement
|
||||
* Add fuzzer for table definitions [#40096](https://github.com/ClickHouse/ClickHouse/pull/40096) ([Anton Popov](https://github.com/CurtizJ)). This represents the biggest advancement for ClickHouse testing in this year so far.
|
||||
|
||||
@ -10,9 +10,11 @@ The following versions of ClickHouse server are currently being supported with s
|
||||
|
||||
| Version | Supported |
|
||||
|:-|:-|
|
||||
| 22.10 | ✔️ |
|
||||
| 22.9 | ✔️ |
|
||||
| 22.8 | ✔️ |
|
||||
| 22.7 | ✔️ |
|
||||
| 22.6 | ✔️ |
|
||||
| 22.7 | ❌ |
|
||||
| 22.6 | ❌ |
|
||||
| 22.5 | ❌ |
|
||||
| 22.4 | ❌ |
|
||||
| 22.3 | ✔️ |
|
||||
|
||||
@ -2,11 +2,11 @@
|
||||
|
||||
# NOTE: has nothing common with DBMS_TCP_PROTOCOL_VERSION,
|
||||
# only DBMS_TCP_PROTOCOL_VERSION should be incremented on protocol changes.
|
||||
SET(VERSION_REVISION 54467)
|
||||
SET(VERSION_REVISION 54468)
|
||||
SET(VERSION_MAJOR 22)
|
||||
SET(VERSION_MINOR 10)
|
||||
SET(VERSION_MINOR 11)
|
||||
SET(VERSION_PATCH 1)
|
||||
SET(VERSION_GITHASH 3030d4c7ff09ec44ab07d0a8069ea923227288a1)
|
||||
SET(VERSION_DESCRIBE v22.10.1.1-testing)
|
||||
SET(VERSION_STRING 22.10.1.1)
|
||||
SET(VERSION_GITHASH 98ab5a3c189232ea2a3dddb9d2be7196ae8b3434)
|
||||
SET(VERSION_DESCRIBE v22.11.1.1-testing)
|
||||
SET(VERSION_STRING 22.11.1.1)
|
||||
# end of autochange
|
||||
|
||||
2
contrib/libcxx
vendored
2
contrib/libcxx
vendored
@ -1 +1 @@
|
||||
Subproject commit 172b2ae074f6755145b91c53a95c8540c1468239
|
||||
Subproject commit 4db7f838afd3139eb3761694b04d31275df45d2d
|
||||
@ -25,6 +25,7 @@ set(SRCS
|
||||
"${LIBCXX_SOURCE_DIR}/src/ios.cpp"
|
||||
"${LIBCXX_SOURCE_DIR}/src/ios.instantiations.cpp"
|
||||
"${LIBCXX_SOURCE_DIR}/src/iostream.cpp"
|
||||
"${LIBCXX_SOURCE_DIR}/src/legacy_debug_handler.cpp"
|
||||
"${LIBCXX_SOURCE_DIR}/src/legacy_pointer_safety.cpp"
|
||||
"${LIBCXX_SOURCE_DIR}/src/locale.cpp"
|
||||
"${LIBCXX_SOURCE_DIR}/src/memory.cpp"
|
||||
@ -49,6 +50,7 @@ set(SRCS
|
||||
"${LIBCXX_SOURCE_DIR}/src/valarray.cpp"
|
||||
"${LIBCXX_SOURCE_DIR}/src/variant.cpp"
|
||||
"${LIBCXX_SOURCE_DIR}/src/vector.cpp"
|
||||
"${LIBCXX_SOURCE_DIR}/src/verbose_abort.cpp"
|
||||
)
|
||||
|
||||
add_library(cxx ${SRCS})
|
||||
|
||||
2
contrib/libcxxabi
vendored
2
contrib/libcxxabi
vendored
@ -1 +1 @@
|
||||
Subproject commit 6eb7cc7a7bdd779e6734d1b9fb451df2274462d7
|
||||
Subproject commit a736a6b3c6a7b8aae2ebad629ca21b2c55b4820e
|
||||
@ -9,6 +9,7 @@ set(SRCS
|
||||
"${LIBCXXABI_SOURCE_DIR}/src/cxa_exception_storage.cpp"
|
||||
"${LIBCXXABI_SOURCE_DIR}/src/cxa_guard.cpp"
|
||||
"${LIBCXXABI_SOURCE_DIR}/src/cxa_handlers.cpp"
|
||||
# "${LIBCXXABI_SOURCE_DIR}/src/cxa_noexception.cpp"
|
||||
"${LIBCXXABI_SOURCE_DIR}/src/cxa_personality.cpp"
|
||||
"${LIBCXXABI_SOURCE_DIR}/src/cxa_thread_atexit.cpp"
|
||||
"${LIBCXXABI_SOURCE_DIR}/src/cxa_vector.cpp"
|
||||
|
||||
2
contrib/rocksdb
vendored
2
contrib/rocksdb
vendored
@ -1 +1 @@
|
||||
Subproject commit e7c2b2f7bcf3b4b33892a1a6d25c32a93edfbdb9
|
||||
Subproject commit 2c8998e26c6d46b27c710d7829c3a15e34959f70
|
||||
2
contrib/zlib-ng
vendored
2
contrib/zlib-ng
vendored
@ -1 +1 @@
|
||||
Subproject commit bffad6f6fe74d6a2f92e2668390664a926c68733
|
||||
Subproject commit 50f0eae1a411764cd6d1e85b3ce471438acd3c1c
|
||||
@ -33,7 +33,7 @@ RUN arch=${TARGETARCH:-amd64} \
|
||||
# lts / testing / prestable / etc
|
||||
ARG REPO_CHANNEL="stable"
|
||||
ARG REPOSITORY="https://packages.clickhouse.com/tgz/${REPO_CHANNEL}"
|
||||
ARG VERSION="22.9.3.18"
|
||||
ARG VERSION="22.10.1.1877"
|
||||
ARG PACKAGES="clickhouse-client clickhouse-server clickhouse-common-static"
|
||||
|
||||
# user/group precreated explicitly with fixed uid/gid on purpose.
|
||||
|
||||
@ -21,7 +21,7 @@ RUN sed -i "s|http://archive.ubuntu.com|${apt_archive}|g" /etc/apt/sources.list
|
||||
|
||||
ARG REPO_CHANNEL="stable"
|
||||
ARG REPOSITORY="deb https://packages.clickhouse.com/deb ${REPO_CHANNEL} main"
|
||||
ARG VERSION="22.9.3.18"
|
||||
ARG VERSION="22.10.1.1877"
|
||||
ARG PACKAGES="clickhouse-client clickhouse-server clickhouse-common-static"
|
||||
|
||||
# set non-empty deb_location_url url to create a docker image
|
||||
|
||||
@ -27,9 +27,14 @@ RUN apt-get update \
|
||||
tar \
|
||||
tzdata \
|
||||
unixodbc \
|
||||
python3-pip \
|
||||
libcurl4-openssl-dev \
|
||||
libssl-dev \
|
||||
&& apt-get clean \
|
||||
&& rm -rf /var/lib/apt/lists/* /var/cache/debconf /tmp/*
|
||||
|
||||
RUN pip3 install pycurl
|
||||
|
||||
# Architecture of the image when BuildKit/buildx is used
|
||||
ARG TARGETARCH
|
||||
|
||||
|
||||
352
docs/changelogs/v22.10.1.1877-stable.md
Normal file
352
docs/changelogs/v22.10.1.1877-stable.md
Normal file
@ -0,0 +1,352 @@
|
||||
---
|
||||
sidebar_position: 1
|
||||
sidebar_label: 2022
|
||||
---
|
||||
|
||||
# 2022 Changelog
|
||||
|
||||
### ClickHouse release v22.10.1.1877-stable (98ab5a3c189) FIXME as compared to v22.9.1.2603-stable (3030d4c7ff0)
|
||||
|
||||
#### Backward Incompatible Change
|
||||
* Rename cache commands: `show caches` -> `show filesystem caches`, `describe cache` -> `describe filesystem cache`. [#41508](https://github.com/ClickHouse/ClickHouse/pull/41508) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Remove support for the `WITH TIMEOUT` section for `LIVE VIEW`. This closes [#40557](https://github.com/ClickHouse/ClickHouse/issues/40557). [#42173](https://github.com/ClickHouse/ClickHouse/pull/42173) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
|
||||
#### New Feature
|
||||
* Add Rust code support into ClickHouse with BLAKE3 hash-function library as an example. [#33435](https://github.com/ClickHouse/ClickHouse/pull/33435) ([BoloniniD](https://github.com/BoloniniD)).
|
||||
* This is the initial implement of Kusto Query Language. (MVP). [#37961](https://github.com/ClickHouse/ClickHouse/pull/37961) ([Yong Wang](https://github.com/kashwy)).
|
||||
* * Support limiting of temporary data stored on disk using settings `max_temporary_data_on_disk_size_for_user`/`max_temporary_data_on_disk_size_for_query` . [#40893](https://github.com/ClickHouse/ClickHouse/pull/40893) ([Vladimir C](https://github.com/vdimir)).
|
||||
* Support Java integers hashing in `javaHash`. [#41131](https://github.com/ClickHouse/ClickHouse/pull/41131) ([JackyWoo](https://github.com/JackyWoo)).
|
||||
* This PR is to support the OpenSSL in-house build like the BoringSSL submodule. Build flag i.e. ENABLE_CH_BUNDLE_BORINGSSL is used to choose between BoringSSL and OpenSSL. By default, the BoringSSL in-house build will be used. [#41142](https://github.com/ClickHouse/ClickHouse/pull/41142) ([MeenaRenganathan22](https://github.com/MeenaRenganathan22)).
|
||||
* Composable protocol configuration is added. [#41198](https://github.com/ClickHouse/ClickHouse/pull/41198) ([Yakov Olkhovskiy](https://github.com/yakov-olkhovskiy)).
|
||||
* Add OpenTelemetry support to ON CLUSTER DDL(require `distributed_ddl_entry_format_version` to be set to 4). [#41484](https://github.com/ClickHouse/ClickHouse/pull/41484) ([Frank Chen](https://github.com/FrankChen021)).
|
||||
* Add setting `format_json_object_each_row_column_for_object_name` to write/parse object name as column value in JSONObjectEachRow format. [#41703](https://github.com/ClickHouse/ClickHouse/pull/41703) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* adds Morton Coding (ZCurve) encode/decode functions. [#41753](https://github.com/ClickHouse/ClickHouse/pull/41753) ([Constantine Peresypkin](https://github.com/pkit)).
|
||||
* Implement support for different UUID binary formats with support for the two most prevalent ones: the default big-endian and Microsoft's mixed-endian as specified in [RFC 4122](https://datatracker.ietf.org/doc/html/rfc4122#section-4.1.1). [#42108](https://github.com/ClickHouse/ClickHouse/pull/42108) ([ltrk2](https://github.com/ltrk2)).
|
||||
* Added an aggregate function `analysisOfVariance` (`anova`) to perform a statistical test over several groups of normally distributed observations to find out whether all groups have the same mean or not. Original PR [#37872](https://github.com/ClickHouse/ClickHouse/issues/37872). [#42131](https://github.com/ClickHouse/ClickHouse/pull/42131) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Add support for `SET setting_name = DEFAULT`. [#42187](https://github.com/ClickHouse/ClickHouse/pull/42187) ([Filatenkov Artur](https://github.com/FArthur-cmd)).
|
||||
* * Add `URL` Functions which conform rfc. Functions include: `cutToFirstSignificantSubdomainCustomRFC`, `cutToFirstSignificantSubdomainCustomWithWWWRFC`, `cutToFirstSignificantSubdomainRFC`, `cutToFirstSignificantSubdomainWithWWWRFC`, `domainRFC`, `domainWithoutWWWRFC`, `firstSignificantSubdomainCustomRFC`, `firstSignificantSubdomainRFC`, `portRFC`, `topLevelDomainRFC`. [#42274](https://github.com/ClickHouse/ClickHouse/pull/42274) ([Quanfa Fu](https://github.com/dentiscalprum)).
|
||||
* Added functions (`randUniform`, `randNormal`, `randLogNormal`, `randExponential`, `randChiSquared`, `randStudentT`, `randFisherF`, `randBernoulli`, `randBinomial`, `randNegativeBinomial`, `randPoisson` ) to generate random values according to the specified distributions. This closes [#21834](https://github.com/ClickHouse/ClickHouse/issues/21834). [#42411](https://github.com/ClickHouse/ClickHouse/pull/42411) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
|
||||
#### Performance Improvement
|
||||
* Implement operator precedence element parser to resolve stack overflow issues and make the required stack size smaller. [#34892](https://github.com/ClickHouse/ClickHouse/pull/34892) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* DISTINCT in order optimization leverage sorting properties of data streams. This improvement will enable reading in order for DISTINCT if applicable (before it was necessary to provide ORDER BY for columns in DISTINCT). [#41014](https://github.com/ClickHouse/ClickHouse/pull/41014) ([Igor Nikonov](https://github.com/devcrafter)).
|
||||
* ColumnVector: optimize UInt8 index with AVX512VBMI. [#41247](https://github.com/ClickHouse/ClickHouse/pull/41247) ([Guo Wangyang](https://github.com/guowangy)).
|
||||
* The performance experiments of **SSB** (Star Schema Benchmark) on the ICX device (Intel Xeon Platinum 8380 CPU, 80 cores, 160 threads) shows that this change could bring a **2.95x** improvement of the geomean of all subcases' QPS. [#41675](https://github.com/ClickHouse/ClickHouse/pull/41675) ([Zhiguo Zhou](https://github.com/ZhiguoZh)).
|
||||
* Fixed slowness in JSONExtract with LowCardinality(String) tuples. [#41726](https://github.com/ClickHouse/ClickHouse/pull/41726) ([AlfVII](https://github.com/AlfVII)).
|
||||
* Add ldapr capabilities to AArch64 builds. This is supported from Graviton 2+, Azure and GCP instances. Only appeared in clang-15 [not so long ago](https://github.com/llvm/llvm-project/commit/9609b5daffe9fd28d83d83da895abc5113f76c24). [#41778](https://github.com/ClickHouse/ClickHouse/pull/41778) ([Daniel Kutenin](https://github.com/danlark1)).
|
||||
* Improve performance when comparing strings and one argument is empty constant string. [#41870](https://github.com/ClickHouse/ClickHouse/pull/41870) ([Jiebin Sun](https://github.com/jiebinn)).
|
||||
* optimize insertFrom of ColumnAggregateFunction to share Aggregate State in some cases. [#41960](https://github.com/ClickHouse/ClickHouse/pull/41960) ([flynn](https://github.com/ucasfl)).
|
||||
* Relax the "Too many parts" threshold. This closes [#6551](https://github.com/ClickHouse/ClickHouse/issues/6551). Now ClickHouse will allow more parts in a partition if the average part size is large enough (at least 10 GiB). This allows to have up to petabytes of data in a single partition of a single table on a single server, which is possible using disk shelves or object storage. [#42002](https://github.com/ClickHouse/ClickHouse/pull/42002) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Make writing to AzureBlobStorage more efficient (respect `max_single_part_upload_size` instead of writing a block per each buffer size). Inefficiency mentioned in [#41754](https://github.com/ClickHouse/ClickHouse/issues/41754). [#42041](https://github.com/ClickHouse/ClickHouse/pull/42041) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Make thread ids in the process list and query_log unique to avoid waste. [#42180](https://github.com/ClickHouse/ClickHouse/pull/42180) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
|
||||
#### Improvement
|
||||
* Added new infrastructure for query analysis and planning under `allow_experimental_analyzer` setting. [#31796](https://github.com/ClickHouse/ClickHouse/pull/31796) ([Maksim Kita](https://github.com/kitaisreal)).
|
||||
* * Support expression `(EXPLAIN SELECT ...)` in a subquery. Queries like `SELECT * FROM (EXPLAIN PIPELINE SELECT col FROM TABLE ORDER BY col)` became valid. [#40630](https://github.com/ClickHouse/ClickHouse/pull/40630) ([Vladimir C](https://github.com/vdimir)).
|
||||
* Currently changing `async_insert_max_data_size` or `async_insert_busy_timeout_ms` in scope of query makes no sense and this leads to bad user experience. E.g. user wants to insert data rarely and he doesn't have an access to server config to tune default settings. [#40668](https://github.com/ClickHouse/ClickHouse/pull/40668) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Embedded Keeper will always start in the background allowing ClickHouse to start without achieving quorum. [#40991](https://github.com/ClickHouse/ClickHouse/pull/40991) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Improvements for reading from remote filesystems, made threadpool size for reads/writes configurable. Closes [#41070](https://github.com/ClickHouse/ClickHouse/issues/41070). [#41011](https://github.com/ClickHouse/ClickHouse/pull/41011) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Made reestablishing a new connection more reactive in case of expiration of the previous one. Previously there was a task which spawns every minute by default and thus a table could be in readonly state for about this time. [#41092](https://github.com/ClickHouse/ClickHouse/pull/41092) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Support all combinators combination in WindowTransform/arratReduce*/initializeAggregation/aggregate functions versioning. Previously combinators like `ForEach/Resample/Map` didn't work in these places, using them led to exception like`State function ... inserts results into non-state column`. [#41107](https://github.com/ClickHouse/ClickHouse/pull/41107) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Now projections can be used with zero copy replication. [#41147](https://github.com/ClickHouse/ClickHouse/pull/41147) ([alesapin](https://github.com/alesapin)).
|
||||
* - Add function tryDecrypt that returns NULL when decrypt fail (e.g. decrypt with incorrect key) instead of throwing exception. [#41206](https://github.com/ClickHouse/ClickHouse/pull/41206) ([Duc Canh Le](https://github.com/canhld94)).
|
||||
* Add the `unreserved_space` column to the `system.disks` table to check how much space is not taken by reservations per disk. [#41254](https://github.com/ClickHouse/ClickHouse/pull/41254) ([filimonov](https://github.com/filimonov)).
|
||||
* Support s3 authorisation headers from ast arguments. [#41261](https://github.com/ClickHouse/ClickHouse/pull/41261) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Add setting 'allow_implicit_no_password' that forbids creating a user with no password unless 'IDENTIFIED WITH no_password' is explicitly specified. [#41341](https://github.com/ClickHouse/ClickHouse/pull/41341) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* keeper-improvement: add support for uploading snapshots to S3. S3 information can be defined inside `keeper_server.s3_snapshot`. [#41342](https://github.com/ClickHouse/ClickHouse/pull/41342) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Add support for MultiRead in Keeper and internal ZooKeeper client. [#41410](https://github.com/ClickHouse/ClickHouse/pull/41410) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* add a support for decimal type comparing with floating point literal in IN operator. [#41544](https://github.com/ClickHouse/ClickHouse/pull/41544) ([liang.huang](https://github.com/lhuang09287750)).
|
||||
* Allow readable size values in cache config. [#41688](https://github.com/ClickHouse/ClickHouse/pull/41688) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Check file path for path traversal attacks in errors logger for input formats. [#41694](https://github.com/ClickHouse/ClickHouse/pull/41694) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* ClickHouse could cache stale DNS entries for some period of time (15 seconds by default) until the cache won't be updated asynchronously. During these period ClickHouse can nevertheless try to establish a connection and produce errors. This behaviour is fixed. [#41707](https://github.com/ClickHouse/ClickHouse/pull/41707) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Add interactive history search with fzf-like utility (fzf/sk) for `clickhouse-client`/`clickhouse-local` (note you can use `FZF_DEFAULT_OPTS`/`SKIM_DEFAULT_OPTIONS` to additionally configure the behavior). [#41730](https://github.com/ClickHouse/ClickHouse/pull/41730) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* For client when connecting to a secure server with invalid certificate only allow to proceed with '--accept-certificate' flag. [#41743](https://github.com/ClickHouse/ClickHouse/pull/41743) ([Yakov Olkhovskiy](https://github.com/yakov-olkhovskiy)).
|
||||
* Add function "tryBase58Decode()", similar to the existing function "tryBase64Decode()". [#41824](https://github.com/ClickHouse/ClickHouse/pull/41824) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Improve feedback when replacing partition with different primary key. Fixes [#34798](https://github.com/ClickHouse/ClickHouse/issues/34798). [#41838](https://github.com/ClickHouse/ClickHouse/pull/41838) ([Salvatore](https://github.com/tbsal)).
|
||||
* Replace back `clickhouse su` command with `sudo -u` in start in order to respect limits in `/etc/security/limits.conf`. [#41847](https://github.com/ClickHouse/ClickHouse/pull/41847) ([Eugene Konkov](https://github.com/ekonkov)).
|
||||
* Fix parallel parsing: segmentator now checks max_block_size. [#41852](https://github.com/ClickHouse/ClickHouse/pull/41852) ([Vitaly Baranov](https://github.com/vitlibar)).
|
||||
* Don't report TABLE_IS_DROPPED exception in order to skip table in case is was just dropped. [#41908](https://github.com/ClickHouse/ClickHouse/pull/41908) ([AlfVII](https://github.com/AlfVII)).
|
||||
* Improve option enable_extended_results_for_datetime_functions to return results of type DateTime64 for functions toStartOfDay, toStartOfHour, toStartOfFifteenMinutes, toStartOfTenMinutes, toStartOfFiveMinutes, toStartOfMinute and timeSlot. [#41910](https://github.com/ClickHouse/ClickHouse/pull/41910) ([Roman Vasin](https://github.com/rvasin)).
|
||||
* Improve DateTime type inference for text formats. Now it respect setting `date_time_input_format` and doesn't try to infer datetimes from numbers as timestamps. Closes [#41389](https://github.com/ClickHouse/ClickHouse/issues/41389) Closes [#42206](https://github.com/ClickHouse/ClickHouse/issues/42206). [#41912](https://github.com/ClickHouse/ClickHouse/pull/41912) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Remove confusing warning when inserting with `perform_ttl_move_on_insert`=false. [#41980](https://github.com/ClickHouse/ClickHouse/pull/41980) ([Vitaly Baranov](https://github.com/vitlibar)).
|
||||
* Allow user to write `countState(*)` similar to `count(*)`. This closes [#9338](https://github.com/ClickHouse/ClickHouse/issues/9338). [#41983](https://github.com/ClickHouse/ClickHouse/pull/41983) ([Amos Bird](https://github.com/amosbird)).
|
||||
* - Fix rankCorr size overflow. [#42020](https://github.com/ClickHouse/ClickHouse/pull/42020) ([Duc Canh Le](https://github.com/canhld94)).
|
||||
* Added an option to specify an arbitrary string as an environment name in the Sentry's config for more handy reports. [#42037](https://github.com/ClickHouse/ClickHouse/pull/42037) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Added system table `asynchronous_insert_log `. It contains information about asynchronous inserts (including results of queries in fire-and-forget mode (with `wait_for_async_insert=0`)) for better introspection. [#42040](https://github.com/ClickHouse/ClickHouse/pull/42040) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Fix parsing out-of-range Date from CSV:. [#42044](https://github.com/ClickHouse/ClickHouse/pull/42044) ([Andrey Zvonov](https://github.com/zvonand)).
|
||||
* parseDataTimeBestEffort support comma between date and time. Closes [#42038](https://github.com/ClickHouse/ClickHouse/issues/42038). [#42049](https://github.com/ClickHouse/ClickHouse/pull/42049) ([flynn](https://github.com/ucasfl)).
|
||||
* Add support for methods lz4, bz2, snappy in 'Accept-Encoding'. [#42071](https://github.com/ClickHouse/ClickHouse/pull/42071) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* Various minor fixes for BLAKE3 function. [#42073](https://github.com/ClickHouse/ClickHouse/pull/42073) ([BoloniniD](https://github.com/BoloniniD)).
|
||||
* Improved stale replica recovery process for `ReplicatedMergeTree`. If lost replica have some parts which absent on a healthy replica, but these parts should appear in future according to replication queue of the healthy replica, then lost replica will keep such parts instead of detaching them. [#42134](https://github.com/ClickHouse/ClickHouse/pull/42134) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Support BACKUP to S3 with as-is path/data structure. [#42232](https://github.com/ClickHouse/ClickHouse/pull/42232) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Add a possibility to use Date32 arguments for date_diff function. Fix issue in date_diff function when using DateTime64 arguments with start date before Unix epoch and end date after Unix epoch. [#42308](https://github.com/ClickHouse/ClickHouse/pull/42308) ([Roman Vasin](https://github.com/rvasin)).
|
||||
* When uploading big parts to minio, 'Complete Multipart Upload' can take a long time. Minio sends heartbeats every 10 seconds (see https://github.com/minio/minio/pull/7198). But clickhouse times out earlier, because the default send/receive timeout is [set](https://github.com/ClickHouse/ClickHouse/blob/cc24fcd6d5dfb67f5f66f5483e986bd1010ad9cf/src/IO/S3/PocoHTTPClient.cpp#L123) to 5 seconds. [#42321](https://github.com/ClickHouse/ClickHouse/pull/42321) ([filimonov](https://github.com/filimonov)).
|
||||
* Add `S3` as a new type of the destination of backups. [#42333](https://github.com/ClickHouse/ClickHouse/pull/42333) ([Vitaly Baranov](https://github.com/vitlibar)).
|
||||
* Fix rarely invalid cast of aggregate state types with complex types such as Decimal. This fixes [#42408](https://github.com/ClickHouse/ClickHouse/issues/42408). [#42417](https://github.com/ClickHouse/ClickHouse/pull/42417) ([Amos Bird](https://github.com/amosbird)).
|
||||
* Support skipping cache completely (both download to cache and reading cached data) in case the requested read range exceeds the threshold defined by cache setting `bypass_cache_threashold`, requires to be enabled with `enable_bypass_cache_with_threshold`). [#42418](https://github.com/ClickHouse/ClickHouse/pull/42418) ([Han Shukai](https://github.com/KinderRiven)).
|
||||
* Merge parts if every part in the range is older than a certain threshold. The threshold can be set by using `min_age_to_force_merge_seconds`. This closes [#35836](https://github.com/ClickHouse/ClickHouse/issues/35836). [#42423](https://github.com/ClickHouse/ClickHouse/pull/42423) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Enabled CompiledExpressionCache in clickhouse-local. [#42477](https://github.com/ClickHouse/ClickHouse/pull/42477) ([AlfVII](https://github.com/AlfVII)).
|
||||
* Remove support for the `{database}` macro from the client's prompt. It was displayed incorrectly if the database was unspecified and it was not updated on `USE` statements. This closes [#25891](https://github.com/ClickHouse/ClickHouse/issues/25891). [#42508](https://github.com/ClickHouse/ClickHouse/pull/42508) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* - Improve the time to recover lost keeper connections. [#42541](https://github.com/ClickHouse/ClickHouse/pull/42541) ([Raúl Marín](https://github.com/Algunenano)).
|
||||
* Allow to use Date32 arguments for dateName function. [#42554](https://github.com/ClickHouse/ClickHouse/pull/42554) ([Roman Vasin](https://github.com/rvasin)).
|
||||
|
||||
#### Bug Fix
|
||||
* Now filters with NULL literals will be used during index analysis. This closes https://github.com/ClickHouse/ClickHouse/pull/41814 [#34063](https://github.com/ClickHouse/ClickHouse/issues/34063). [#41842](https://github.com/ClickHouse/ClickHouse/pull/41842) ([Amos Bird](https://github.com/amosbird)).
|
||||
* - Choose correct aggregation method for LowCardinality with BigInt. [#42342](https://github.com/ClickHouse/ClickHouse/pull/42342) ([Duc Canh Le](https://github.com/canhld94)).
|
||||
* Fix using subqueries in row policy filters. This PR fixes [#32463](https://github.com/ClickHouse/ClickHouse/issues/32463). [#42562](https://github.com/ClickHouse/ClickHouse/pull/42562) ([Vitaly Baranov](https://github.com/vitlibar)).
|
||||
|
||||
#### Build/Testing/Packaging Improvement
|
||||
* Added support of WHERE clause generation to AST Fuzzer and possibility to add or remove ORDER BY and WHERE clause. [#38519](https://github.com/ClickHouse/ClickHouse/pull/38519) ([Ilya Yatsishin](https://github.com/qoega)).
|
||||
* Aarch64 binaries now require at least ARMv8.2, released in 2016. Most notably, this enables use of ARM LSE, i.e. native atomic operations. Also, CMake build option "NO_ARMV81_OR_HIGHER" has been added to allow compilation of binaries for older ARMv8.0 hardware, e.g. Raspberry Pi 4. [#41610](https://github.com/ClickHouse/ClickHouse/pull/41610) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* After updating runners to 22.04 cgroups stopped to work in privileged mode, here's the issue https://github.com/moby/moby/issues/42275#issuecomment-1115055846. [#41857](https://github.com/ClickHouse/ClickHouse/pull/41857) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Allow building ClickHouse with Musl (small changes after it was already supported but broken). [#41987](https://github.com/ClickHouse/ClickHouse/pull/41987) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* - Add the `$CLICKHOUSE_CRONFILE` file checking to avoid running the `sed` command to get the file not found error. [#42081](https://github.com/ClickHouse/ClickHouse/pull/42081) ([Chun-Sheng, Li](https://github.com/peter279k)).
|
||||
* Update cctz to the latest master, update tzdb to 2020e. [#42273](https://github.com/ClickHouse/ClickHouse/pull/42273) ([Dom Del Nano](https://github.com/ddelnano)).
|
||||
* Update tzdata to 2022e to support the new timezone changes. Palestine transitions are now Saturdays at 02:00. Simplify three Ukraine zones into one. Jordan and Syria switch from +02/+03 with DST to year-round +03. (https://data.iana.org/time-zones/tzdb/NEWS). This closes [#42252](https://github.com/ClickHouse/ClickHouse/issues/42252). [#42327](https://github.com/ClickHouse/ClickHouse/pull/42327) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Fix power8 support. [#42462](https://github.com/ClickHouse/ClickHouse/pull/42462) ([Boris Kuschel](https://github.com/bkuschel)).
|
||||
|
||||
#### Bug Fix (user-visible misbehavior in official stable or prestable release)
|
||||
|
||||
* Several fixes for DiskWeb. [#41652](https://github.com/ClickHouse/ClickHouse/pull/41652) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Fixes issue when docker run will fail if "https_port" is not present in config. [#41693](https://github.com/ClickHouse/ClickHouse/pull/41693) ([Yakov Olkhovskiy](https://github.com/yakov-olkhovskiy)).
|
||||
* Mutations were not cancelled properly on server shutdown or `SYSTEM STOP MERGES` query and cancellation might take long time, it's fixed. [#41699](https://github.com/ClickHouse/ClickHouse/pull/41699) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Fix wrong result of queries with `ORDER BY` or `GROUP BY` by columns from prefix of sorting key, wrapped into monotonic functions, with enable "read in order" optimization (settings `optimize_read_in_order` and `optimize_aggregation_in_order`). [#41701](https://github.com/ClickHouse/ClickHouse/pull/41701) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Fix possible crash in `SELECT` from `Merge` table with enabled `optimize_monotonous_functions_in_order_by` setting. Fixes [#41269](https://github.com/ClickHouse/ClickHouse/issues/41269). [#41740](https://github.com/ClickHouse/ClickHouse/pull/41740) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Fixed "Part ... intersects part ..." error that might happen in extremely rare cases if replica was restarted just after detaching some part as broken. [#41741](https://github.com/ClickHouse/ClickHouse/pull/41741) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Don't allow to create or alter merge tree tables with virtual column name _row_exists, which is reserved for lightweight delete. Fixed [#41716](https://github.com/ClickHouse/ClickHouse/issues/41716). [#41763](https://github.com/ClickHouse/ClickHouse/pull/41763) ([Jianmei Zhang](https://github.com/zhangjmruc)).
|
||||
* Fix a bug that CORS headers are missing in some HTTP responses. [#41792](https://github.com/ClickHouse/ClickHouse/pull/41792) ([Frank Chen](https://github.com/FrankChen021)).
|
||||
* 22.9 might fail to startup `ReplicatedMergeTree` table if that table was created by 20.3 or older version and was never altered, it's fixed. Fixes [#41742](https://github.com/ClickHouse/ClickHouse/issues/41742). [#41796](https://github.com/ClickHouse/ClickHouse/pull/41796) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* When the batch sending fails for some reason, it cannot be automatically recovered, and if it is not processed in time, it will lead to accumulation, and the printed error message will become longer and longer, which will cause the http thread to block. [#41813](https://github.com/ClickHouse/ClickHouse/pull/41813) ([zhongyuankai](https://github.com/zhongyuankai)).
|
||||
* Fix compact parts with compressed marks setting. Fixes [#41783](https://github.com/ClickHouse/ClickHouse/issues/41783) and [#41746](https://github.com/ClickHouse/ClickHouse/issues/41746). [#41823](https://github.com/ClickHouse/ClickHouse/pull/41823) ([alesapin](https://github.com/alesapin)).
|
||||
* Old versions of Replicated database doesn't have a special marker in [Zoo]Keeper. We need to check only whether the node contains come obscure data instead of special mark. [#41875](https://github.com/ClickHouse/ClickHouse/pull/41875) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Fix possible exception in fs cache. [#41884](https://github.com/ClickHouse/ClickHouse/pull/41884) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Fix use_environment_credentials for s3 table function. [#41970](https://github.com/ClickHouse/ClickHouse/pull/41970) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Fixed "Directory already exists and is not empty" error on detaching broken part that might prevent `ReplicatedMergeTree` table from starting replication. Fixes [#40957](https://github.com/ClickHouse/ClickHouse/issues/40957). [#41981](https://github.com/ClickHouse/ClickHouse/pull/41981) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* toDateTime64() now returns the same output with negative integer and float arguments. [#42025](https://github.com/ClickHouse/ClickHouse/pull/42025) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Fix write into AzureBlobStorage. Partially closes [#41754](https://github.com/ClickHouse/ClickHouse/issues/41754). [#42034](https://github.com/ClickHouse/ClickHouse/pull/42034) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Fix the bzip2 decoding issue for specific bzip2 files. [#42046](https://github.com/ClickHouse/ClickHouse/pull/42046) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* - Fix SQL function "toLastDayOfMonth()" with setting "enable_extended_results_for_datetime_functions = 1" at the beginning of the extended range (January 1900). - Fix SQL function "toRelativeWeekNum()" with setting "enable_extended_results_for_datetime_functions = 1" at the end of extended range (December 2299). - Improve the performance of for SQL functions "toISOYear()", "toFirstDayNumOfISOYearIndex()" and "toYearWeekOfNewyearMode()" by avoiding unnecessary index arithmetics. [#42084](https://github.com/ClickHouse/ClickHouse/pull/42084) ([Roman Vasin](https://github.com/rvasin)).
|
||||
* The maximum size of fetches for each table accidentally was set to 8 while the pool size could be bigger. Now the maximum size of fetches for table is equal to the pool size. [#42090](https://github.com/ClickHouse/ClickHouse/pull/42090) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* A table might be shut down and a dictionary might be detached before checking if can be dropped without breaking dependencies between table, it's fixed. Fixes [#41982](https://github.com/ClickHouse/ClickHouse/issues/41982). [#42106](https://github.com/ClickHouse/ClickHouse/pull/42106) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Fix bad inefficiency of `remote_filesystem_read_method=read` with filesystem cache. Closes [#42125](https://github.com/ClickHouse/ClickHouse/issues/42125). [#42129](https://github.com/ClickHouse/ClickHouse/pull/42129) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Fix possible timeout exception for distributed queries with use_hedged_requests=0. [#42130](https://github.com/ClickHouse/ClickHouse/pull/42130) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Fixed a minor bug inside function `runningDifference` in case of using it with `Date32` type. Previously `Date` was used and it may cause some logical errors like `Bad cast from type DB::ColumnVector<int> to DB::ColumnVector<unsigned short>'`. [#42143](https://github.com/ClickHouse/ClickHouse/pull/42143) ([Alfred Xu](https://github.com/sperlingxx)).
|
||||
* Fix reusing of files > 4GB from base backup. [#42146](https://github.com/ClickHouse/ClickHouse/pull/42146) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* DISTINCT in order fails with LOGICAL_ERROR if first column in sorting key contains function. [#42186](https://github.com/ClickHouse/ClickHouse/pull/42186) ([Igor Nikonov](https://github.com/devcrafter)).
|
||||
* Fix a bug with projections and the `aggregate_functions_null_for_empty` setting. This bug is very rare and appears only if you enable the `aggregate_functions_null_for_empty` setting in the server's config. This closes [#41647](https://github.com/ClickHouse/ClickHouse/issues/41647). [#42198](https://github.com/ClickHouse/ClickHouse/pull/42198) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* - Fix read from buffer with read in order desc. [#42236](https://github.com/ClickHouse/ClickHouse/pull/42236) ([Duc Canh Le](https://github.com/canhld94)).
|
||||
* Fix a bug which prevents ClickHouse to start when background_pool_size setting is set on default profile but background_merges_mutations_concurrency_ratio is not. [#42315](https://github.com/ClickHouse/ClickHouse/pull/42315) ([nvartolomei](https://github.com/nvartolomei)).
|
||||
* `ALTER UPDATE` of attached part (with columns different from table schema) could create an invalid `columns.txt` metadata on disk. Reading from such part could fail with errors or return invalid data. Fixes [#42161](https://github.com/ClickHouse/ClickHouse/issues/42161). [#42319](https://github.com/ClickHouse/ClickHouse/pull/42319) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Setting `additional_table_filters` were not applied to `Distributed` storage. Fixes [#41692](https://github.com/ClickHouse/ClickHouse/issues/41692). [#42322](https://github.com/ClickHouse/ClickHouse/pull/42322) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Fix a data race in query finish/cancel. This closes [#42346](https://github.com/ClickHouse/ClickHouse/issues/42346). [#42362](https://github.com/ClickHouse/ClickHouse/pull/42362) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* This reverts [#40217](https://github.com/ClickHouse/ClickHouse/issues/40217) which introduced a regression in date/time functions. [#42367](https://github.com/ClickHouse/ClickHouse/pull/42367) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Fix assert cast in join on falsy condition, Close [#42380](https://github.com/ClickHouse/ClickHouse/issues/42380). [#42407](https://github.com/ClickHouse/ClickHouse/pull/42407) ([Vladimir C](https://github.com/vdimir)).
|
||||
* Fix buffer overflow in the processing of Decimal data types. This closes [#42451](https://github.com/ClickHouse/ClickHouse/issues/42451). [#42465](https://github.com/ClickHouse/ClickHouse/pull/42465) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* `AggregateFunctionQuantile` now correctly works with UInt128 columns. Previously, the quantile state interpreted `UInt128` columns as `Int128` which could have led to incorrect results. [#42473](https://github.com/ClickHouse/ClickHouse/pull/42473) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Fix bad_assert during INSERT into Annoy indexes over non-Float32 columns. [#42485](https://github.com/ClickHouse/ClickHouse/pull/42485) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* This closes [#42453](https://github.com/ClickHouse/ClickHouse/issues/42453). [#42573](https://github.com/ClickHouse/ClickHouse/pull/42573) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Fix function `arrayElement` with type `Map` with `Nullable` values and `Nullable` index. [#42623](https://github.com/ClickHouse/ClickHouse/pull/42623) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
|
||||
#### Bug Fix (user-visible misbehaviour in official stable or prestable release)
|
||||
|
||||
* Fix unexpected table loading error when partition key contains alias function names during server upgrade. [#36379](https://github.com/ClickHouse/ClickHouse/pull/36379) ([Amos Bird](https://github.com/amosbird)).
|
||||
|
||||
#### Build Improvement
|
||||
|
||||
* Fixed SipHash Endian issue for s390x platform. [#41372](https://github.com/ClickHouse/ClickHouse/pull/41372) ([Harry Lee](https://github.com/HarryLeeIBM)).
|
||||
* Enable lib base64 for ppc64le platform. [#41974](https://github.com/ClickHouse/ClickHouse/pull/41974) ([Suzy Wang](https://github.com/SuzyWangIBMer)).
|
||||
* Fixed Endian issue in T64 compression codec on s390x. [#42314](https://github.com/ClickHouse/ClickHouse/pull/42314) ([Harry Lee](https://github.com/HarryLeeIBM)).
|
||||
|
||||
#### NO CL ENTRY
|
||||
|
||||
* NO CL ENTRY: 'Revert "Disable parallel s3 multipart upload for part moves."'. [#41681](https://github.com/ClickHouse/ClickHouse/pull/41681) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* NO CL ENTRY: 'Revert "Attempt to fix abort from parallel parsing"'. [#42545](https://github.com/ClickHouse/ClickHouse/pull/42545) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* NO CL ENTRY: 'Revert "Low cardinality cases moved to the function for its corresponding type"'. [#42633](https://github.com/ClickHouse/ClickHouse/pull/42633) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
|
||||
#### NOT FOR CHANGELOG / INSIGNIFICANT
|
||||
|
||||
* Test for ignore function in PARTITION KEY [#39875](https://github.com/ClickHouse/ClickHouse/pull/39875) ([UnamedRus](https://github.com/UnamedRus)).
|
||||
* Add fuzzer for table definitions [#40096](https://github.com/ClickHouse/ClickHouse/pull/40096) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Add missing tests for legacy geobase [#40684](https://github.com/ClickHouse/ClickHouse/pull/40684) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Remove obsolete comment from the config.xml [#41518](https://github.com/ClickHouse/ClickHouse/pull/41518) ([filimonov](https://github.com/filimonov)).
|
||||
* Resurrect parallel distributed insert select with s3Cluster [#41535](https://github.com/ClickHouse/ClickHouse/pull/41535) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Update runners to a recent version to install on 22.04 [#41556](https://github.com/ClickHouse/ClickHouse/pull/41556) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Refactor wiping sensitive information from logs. [#41562](https://github.com/ClickHouse/ClickHouse/pull/41562) ([Vitaly Baranov](https://github.com/vitlibar)).
|
||||
* Better S3 logs [#41587](https://github.com/ClickHouse/ClickHouse/pull/41587) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* Fix typos in JSON formats after [#40910](https://github.com/ClickHouse/ClickHouse/issues/40910) [#41614](https://github.com/ClickHouse/ClickHouse/pull/41614) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Fix drop for KeeperMap [#41616](https://github.com/ClickHouse/ClickHouse/pull/41616) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* increase default max_suspicious_broken_parts to 100 [#41619](https://github.com/ClickHouse/ClickHouse/pull/41619) ([Denny Crane](https://github.com/den-crane)).
|
||||
* Release AWS SDK log level + replace one exception [#41649](https://github.com/ClickHouse/ClickHouse/pull/41649) ([alesapin](https://github.com/alesapin)).
|
||||
* Fix a destruction order for views ThreadStatus [#41650](https://github.com/ClickHouse/ClickHouse/pull/41650) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Add very explicit logging on disk choice for fetch [#41653](https://github.com/ClickHouse/ClickHouse/pull/41653) ([alesapin](https://github.com/alesapin)).
|
||||
* Fix race between ~BackgroundSchedulePool and ~DNSCacheUpdater [#41654](https://github.com/ClickHouse/ClickHouse/pull/41654) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Add changelog for 22.9 [#41668](https://github.com/ClickHouse/ClickHouse/pull/41668) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Update version after release [#41670](https://github.com/ClickHouse/ClickHouse/pull/41670) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Fix error message [#41680](https://github.com/ClickHouse/ClickHouse/pull/41680) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Add test for setting output_format_json_validate_utf8 [#41691](https://github.com/ClickHouse/ClickHouse/pull/41691) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Resolve findings from clang-tidy [#41702](https://github.com/ClickHouse/ClickHouse/pull/41702) ([ltrk2](https://github.com/ltrk2)).
|
||||
* Ignore Keeper errors from ReplicatedMergeTreeAttachThread in stress tests [#41717](https://github.com/ClickHouse/ClickHouse/pull/41717) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Collect logs in Stress test using clickhouse-local [#41721](https://github.com/ClickHouse/ClickHouse/pull/41721) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Disable flaky `test_merge_tree_azure_blob_storage` [#41722](https://github.com/ClickHouse/ClickHouse/pull/41722) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Update version_date.tsv and changelogs after v22.9.2.7-stable [#41724](https://github.com/ClickHouse/ClickHouse/pull/41724) ([robot-clickhouse](https://github.com/robot-clickhouse)).
|
||||
* Fix part removal retries [#41728](https://github.com/ClickHouse/ClickHouse/pull/41728) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Try fix azure tests [#41731](https://github.com/ClickHouse/ClickHouse/pull/41731) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Fix test build [#41732](https://github.com/ClickHouse/ClickHouse/pull/41732) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Change logging levels in cache [#41733](https://github.com/ClickHouse/ClickHouse/pull/41733) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Revert of "Revert the revert of "ColumnVector: optimize filter with AVX512 VBMI2 compress store" [#40033](https://github.com/ClickHouse/ClickHouse/issues/40033)" [#41752](https://github.com/ClickHouse/ClickHouse/pull/41752) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Fix SET query parameters formatting [#41755](https://github.com/ClickHouse/ClickHouse/pull/41755) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* Support to run testcases on macOS [#41760](https://github.com/ClickHouse/ClickHouse/pull/41760) ([Frank Chen](https://github.com/FrankChen021)).
|
||||
* Bump LLVM from 12 to 13 [#41762](https://github.com/ClickHouse/ClickHouse/pull/41762) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* ColumnVector: re-enable AVX512_VBMI/AVX512_VBMI2 optimized filter and index [#41765](https://github.com/ClickHouse/ClickHouse/pull/41765) ([Guo Wangyang](https://github.com/guowangy)).
|
||||
* Update 02354_annoy.sql [#41767](https://github.com/ClickHouse/ClickHouse/pull/41767) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Fix the typo preventing building latest images [#41769](https://github.com/ClickHouse/ClickHouse/pull/41769) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Make automatic download script choose between ARMv8.0 or ARMv8.2 builds [#41775](https://github.com/ClickHouse/ClickHouse/pull/41775) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Fix tests for docker-ci [#41777](https://github.com/ClickHouse/ClickHouse/pull/41777) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Possible fix for KeeperMap drop [#41784](https://github.com/ClickHouse/ClickHouse/pull/41784) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Fix drop of completely dropped table [#41789](https://github.com/ClickHouse/ClickHouse/pull/41789) ([alesapin](https://github.com/alesapin)).
|
||||
* Log git hash during startup [#41790](https://github.com/ClickHouse/ClickHouse/pull/41790) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Revert "ColumnVector: optimize UInt8 index with AVX512VBMI ([#41247](https://github.com/ClickHouse/ClickHouse/issues/41247))" [#41797](https://github.com/ClickHouse/ClickHouse/pull/41797) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Small fix in dashboard [#41798](https://github.com/ClickHouse/ClickHouse/pull/41798) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Keep the most important log in stress tests [#41821](https://github.com/ClickHouse/ClickHouse/pull/41821) ([alesapin](https://github.com/alesapin)).
|
||||
* Use copy for some operations instead of hardlinks [#41832](https://github.com/ClickHouse/ClickHouse/pull/41832) ([alesapin](https://github.com/alesapin)).
|
||||
* Remove unused variable in registerStorageMergeTree.cpp [#41839](https://github.com/ClickHouse/ClickHouse/pull/41839) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Fix Jepsen [#41845](https://github.com/ClickHouse/ClickHouse/pull/41845) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Increase `request_timeout_ms` for s3 tests in CI [#41853](https://github.com/ClickHouse/ClickHouse/pull/41853) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* tests: fix debug symbols (and possible crashes) for backward compatiblity check [#41854](https://github.com/ClickHouse/ClickHouse/pull/41854) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Remove two redundant lines [#41856](https://github.com/ClickHouse/ClickHouse/pull/41856) ([alesapin](https://github.com/alesapin)).
|
||||
* Infer Object type only when allow_experimental_object_type is enabled [#41858](https://github.com/ClickHouse/ClickHouse/pull/41858) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Add default UNION/EXCEPT/INTERSECT to the echo query text [#41862](https://github.com/ClickHouse/ClickHouse/pull/41862) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* Consolidate CMake-generated config headers [#41873](https://github.com/ClickHouse/ClickHouse/pull/41873) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Fix 02267_file_globs_schema_inference.sql flakiness [#41877](https://github.com/ClickHouse/ClickHouse/pull/41877) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Docs: Remove obsolete modelEvaluate() mention [#41878](https://github.com/ClickHouse/ClickHouse/pull/41878) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Better exception message for duplicate column names in schema inference [#41885](https://github.com/ClickHouse/ClickHouse/pull/41885) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Docs: Reference external papers as DOIs [#41886](https://github.com/ClickHouse/ClickHouse/pull/41886) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Make LDAPR a prerequisite for downloading the ARMv8.2 build [#41897](https://github.com/ClickHouse/ClickHouse/pull/41897) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Another sync replicas in test_recovery_replica [#41898](https://github.com/ClickHouse/ClickHouse/pull/41898) ([Yakov Olkhovskiy](https://github.com/yakov-olkhovskiy)).
|
||||
* remove unused code [#41921](https://github.com/ClickHouse/ClickHouse/pull/41921) ([flynn](https://github.com/ucasfl)).
|
||||
* Move all queries for MV creation to the end of queue during recovering [#41932](https://github.com/ClickHouse/ClickHouse/pull/41932) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Fix broken test_disks_app_func [#41933](https://github.com/ClickHouse/ClickHouse/pull/41933) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Temporarily disable ThreadFuzzer with TSan [#41943](https://github.com/ClickHouse/ClickHouse/pull/41943) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Enable some disabled S3 tests [#41945](https://github.com/ClickHouse/ClickHouse/pull/41945) ([alesapin](https://github.com/alesapin)).
|
||||
* QOL log improvements [#41947](https://github.com/ClickHouse/ClickHouse/pull/41947) ([Raúl Marín](https://github.com/Algunenano)).
|
||||
* Fix non-deterministic test results [#41948](https://github.com/ClickHouse/ClickHouse/pull/41948) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Earlier throw exception in PullingAsyncPipelineExecutor. [#41949](https://github.com/ClickHouse/ClickHouse/pull/41949) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Fix linker error [#41950](https://github.com/ClickHouse/ClickHouse/pull/41950) ([ltrk2](https://github.com/ltrk2)).
|
||||
* Bump LLVM from 13 to 14 [#41951](https://github.com/ClickHouse/ClickHouse/pull/41951) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Update version_date.tsv and changelogs after v22.3.13.80-lts [#41953](https://github.com/ClickHouse/ClickHouse/pull/41953) ([robot-clickhouse](https://github.com/robot-clickhouse)).
|
||||
* Update version_date.tsv and changelogs after v22.7.6.74-stable [#41954](https://github.com/ClickHouse/ClickHouse/pull/41954) ([robot-clickhouse](https://github.com/robot-clickhouse)).
|
||||
* Update version_date.tsv and changelogs after v22.8.6.71-lts [#41955](https://github.com/ClickHouse/ClickHouse/pull/41955) ([robot-clickhouse](https://github.com/robot-clickhouse)).
|
||||
* Update version_date.tsv and changelogs after v22.9.3.18-stable [#41956](https://github.com/ClickHouse/ClickHouse/pull/41956) ([robot-clickhouse](https://github.com/robot-clickhouse)).
|
||||
* Add a warning message to release.py script, require release type [#41975](https://github.com/ClickHouse/ClickHouse/pull/41975) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Rename max_temp_data_on_disk -> max_temporary_data_on_disk [#41984](https://github.com/ClickHouse/ClickHouse/pull/41984) ([Vladimir C](https://github.com/vdimir)).
|
||||
* Add more checkStackSize calls [#41991](https://github.com/ClickHouse/ClickHouse/pull/41991) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Fix test 02403_big_http_chunk_size [#41996](https://github.com/ClickHouse/ClickHouse/pull/41996) ([Vitaly Baranov](https://github.com/vitlibar)).
|
||||
* More sane behavior of part number thresholds override in query level settings [#42001](https://github.com/ClickHouse/ClickHouse/pull/42001) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Remove useless code [#42004](https://github.com/ClickHouse/ClickHouse/pull/42004) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Refactoring: Uninline some error handling methods [#42010](https://github.com/ClickHouse/ClickHouse/pull/42010) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Fix warning that ENABLE_REPLXX is unused [#42013](https://github.com/ClickHouse/ClickHouse/pull/42013) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Drop leftovers of libexecinfo [#42014](https://github.com/ClickHouse/ClickHouse/pull/42014) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* More detailed exception message [#42022](https://github.com/ClickHouse/ClickHouse/pull/42022) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Build against an LLVM version which has clang[-extra-tools], lldb and lld removed [#42023](https://github.com/ClickHouse/ClickHouse/pull/42023) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Add log message and lower the retry timeout in MergeTreeRestartingThread [#42026](https://github.com/ClickHouse/ClickHouse/pull/42026) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Update amqp-cpp [#42031](https://github.com/ClickHouse/ClickHouse/pull/42031) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Fix No such key during table drop [#42036](https://github.com/ClickHouse/ClickHouse/pull/42036) ([alesapin](https://github.com/alesapin)).
|
||||
* Temporarily disable too aggressive tests [#42050](https://github.com/ClickHouse/ClickHouse/pull/42050) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Fix style check [#42055](https://github.com/ClickHouse/ClickHouse/pull/42055) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Function name normalization fix functions header [#42063](https://github.com/ClickHouse/ClickHouse/pull/42063) ([Maksim Kita](https://github.com/kitaisreal)).
|
||||
* remove unused virtual keyword [#42065](https://github.com/ClickHouse/ClickHouse/pull/42065) ([flynn](https://github.com/ucasfl)).
|
||||
* Fix crash in `SummingMergeTree` with `LowCardinality` [#42066](https://github.com/ClickHouse/ClickHouse/pull/42066) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Fix drop of completely dropped table [#42067](https://github.com/ClickHouse/ClickHouse/pull/42067) ([alesapin](https://github.com/alesapin)).
|
||||
* Fix assertion in bloom filter index [#42072](https://github.com/ClickHouse/ClickHouse/pull/42072) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Ignore core.autocrlf for tests references [#42076](https://github.com/ClickHouse/ClickHouse/pull/42076) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Fix progress for INSERT SELECT [#42078](https://github.com/ClickHouse/ClickHouse/pull/42078) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Avoid adding extra new line after using fuzzy history search [#42080](https://github.com/ClickHouse/ClickHouse/pull/42080) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Add `at` to runner AMI, bump gh runner version [#42082](https://github.com/ClickHouse/ClickHouse/pull/42082) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Use send_metadata instead of send_object_metadata [#42085](https://github.com/ClickHouse/ClickHouse/pull/42085) ([Elena Torró](https://github.com/elenatorro)).
|
||||
* Docs: Preparations to remove misc statements page [#42086](https://github.com/ClickHouse/ClickHouse/pull/42086) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Followup for TemporaryDataOnDisk [#42103](https://github.com/ClickHouse/ClickHouse/pull/42103) ([Vladimir C](https://github.com/vdimir)).
|
||||
* Disable 02122_join_group_by_timeout for debug [#42104](https://github.com/ClickHouse/ClickHouse/pull/42104) ([Vladimir C](https://github.com/vdimir)).
|
||||
* Update version_date.tsv and changelogs after v22.6.9.11-stable [#42114](https://github.com/ClickHouse/ClickHouse/pull/42114) ([robot-clickhouse](https://github.com/robot-clickhouse)).
|
||||
* JIT compilation migration to LLVM 15 [#42123](https://github.com/ClickHouse/ClickHouse/pull/42123) ([Maksim Kita](https://github.com/kitaisreal)).
|
||||
* Fix build without TSA [#42128](https://github.com/ClickHouse/ClickHouse/pull/42128) ([Raúl Marín](https://github.com/Algunenano)).
|
||||
* Update codespell-ignore-words.list [#42132](https://github.com/ClickHouse/ClickHouse/pull/42132) ([Dan Roscigno](https://github.com/DanRoscigno)).
|
||||
* Add null pointer checks [#42135](https://github.com/ClickHouse/ClickHouse/pull/42135) ([ltrk2](https://github.com/ltrk2)).
|
||||
* Revert [#27787](https://github.com/ClickHouse/ClickHouse/issues/27787) [#42136](https://github.com/ClickHouse/ClickHouse/pull/42136) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Follow up for [#42129](https://github.com/ClickHouse/ClickHouse/issues/42129) [#42144](https://github.com/ClickHouse/ClickHouse/pull/42144) ([Kseniia Sumarokova](https://github.com/kssenii)).
|
||||
* Fix checking parent for old-format parts [#42147](https://github.com/ClickHouse/ClickHouse/pull/42147) ([alesapin](https://github.com/alesapin)).
|
||||
* Revert "Resurrect parallel distributed insert select with s3Cluster [#42150](https://github.com/ClickHouse/ClickHouse/pull/42150) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Docs: Add "TABLE" to CHECK/DESCRIBE statements in sidebar [#42152](https://github.com/ClickHouse/ClickHouse/pull/42152) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Add logging during merge tree startup [#42163](https://github.com/ClickHouse/ClickHouse/pull/42163) ([alesapin](https://github.com/alesapin)).
|
||||
* Abort instead of `__builtin_unreachable` in debug builds [#42168](https://github.com/ClickHouse/ClickHouse/pull/42168) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* [RFC] Enable -Wshorten-64-to-32 [#42190](https://github.com/ClickHouse/ClickHouse/pull/42190) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Fix dialect setting description [#42196](https://github.com/ClickHouse/ClickHouse/pull/42196) ([Yakov Olkhovskiy](https://github.com/yakov-olkhovskiy)).
|
||||
* Add a test for #658 [#42197](https://github.com/ClickHouse/ClickHouse/pull/42197) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* use alias for MergeMutateSelectedEntry share ptr [#42211](https://github.com/ClickHouse/ClickHouse/pull/42211) ([Tian Xinhui](https://github.com/xinhuitian)).
|
||||
* Fix LLVM build [#42216](https://github.com/ClickHouse/ClickHouse/pull/42216) ([Raúl Marín](https://github.com/Algunenano)).
|
||||
* Exclude comments from style-check defined extern [#42217](https://github.com/ClickHouse/ClickHouse/pull/42217) ([Vladimir C](https://github.com/vdimir)).
|
||||
* Update safeExit.cpp [#42220](https://github.com/ClickHouse/ClickHouse/pull/42220) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Disable concurrent parts removal [#42222](https://github.com/ClickHouse/ClickHouse/pull/42222) ([alesapin](https://github.com/alesapin)).
|
||||
* Fail fast on empty URL in HDFS [#42223](https://github.com/ClickHouse/ClickHouse/pull/42223) ([Ilya Yatsishin](https://github.com/qoega)).
|
||||
* Add a test for [#2389](https://github.com/ClickHouse/ClickHouse/issues/2389) [#42235](https://github.com/ClickHouse/ClickHouse/pull/42235) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Use MultiRead where possible [#42243](https://github.com/ClickHouse/ClickHouse/pull/42243) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Minor cleanups of LLVM integration [#42249](https://github.com/ClickHouse/ClickHouse/pull/42249) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* remove useless code [#42253](https://github.com/ClickHouse/ClickHouse/pull/42253) ([flynn](https://github.com/ucasfl)).
|
||||
* Early return of corner cases in selectPartsToMutate function [#42254](https://github.com/ClickHouse/ClickHouse/pull/42254) ([Tian Xinhui](https://github.com/xinhuitian)).
|
||||
* Refactor the implementation of user-defined functions [#42263](https://github.com/ClickHouse/ClickHouse/pull/42263) ([Vitaly Baranov](https://github.com/vitlibar)).
|
||||
* assert unused value in test_replicated_merge_tree_compatibility [#42266](https://github.com/ClickHouse/ClickHouse/pull/42266) ([nvartolomei](https://github.com/nvartolomei)).
|
||||
* Fix Date Interval add/minus over DataTypeDate32 [#42279](https://github.com/ClickHouse/ClickHouse/pull/42279) ([Alfred Xu](https://github.com/sperlingxx)).
|
||||
* Fix log-level in `clickhouse-disks` [#42302](https://github.com/ClickHouse/ClickHouse/pull/42302) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* Remove forgotten debug logging [#42313](https://github.com/ClickHouse/ClickHouse/pull/42313) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Fix another trash in zero-copy replication [#42317](https://github.com/ClickHouse/ClickHouse/pull/42317) ([alesapin](https://github.com/alesapin)).
|
||||
* go update for diagnostics tool [#42325](https://github.com/ClickHouse/ClickHouse/pull/42325) ([Dale McDiarmid](https://github.com/gingerwizard)).
|
||||
* Better logging for asynchronous inserts [#42345](https://github.com/ClickHouse/ClickHouse/pull/42345) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Use nfpm packager for archlinux packages [#42349](https://github.com/ClickHouse/ClickHouse/pull/42349) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Bump llvm/clang to 15.0.2 [#42351](https://github.com/ClickHouse/ClickHouse/pull/42351) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Make getResource() independent from the order of the sections [#42353](https://github.com/ClickHouse/ClickHouse/pull/42353) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Smaller threshold for multipart upload part size increase [#42392](https://github.com/ClickHouse/ClickHouse/pull/42392) ([alesapin](https://github.com/alesapin)).
|
||||
* Better error message for unsupported delimiters in custom formats [#42406](https://github.com/ClickHouse/ClickHouse/pull/42406) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Fix formatting of `ALTER FREEZE` [#42409](https://github.com/ClickHouse/ClickHouse/pull/42409) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Replace table name in ast fuzzer more often [#42413](https://github.com/ClickHouse/ClickHouse/pull/42413) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Add *-15 tools to cmake.tools for GCC build [#42430](https://github.com/ClickHouse/ClickHouse/pull/42430) ([Ilya Yatsishin](https://github.com/qoega)).
|
||||
* Deactivate tasks in ReplicatedMergeTree until startup [#42441](https://github.com/ClickHouse/ClickHouse/pull/42441) ([alesapin](https://github.com/alesapin)).
|
||||
* Revert "Revert [#27787](https://github.com/ClickHouse/ClickHouse/issues/27787)" [#42442](https://github.com/ClickHouse/ClickHouse/pull/42442) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Update woboq_codebrowser location [#42448](https://github.com/ClickHouse/ClickHouse/pull/42448) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* add mdx and jsx to list of doc files [#42454](https://github.com/ClickHouse/ClickHouse/pull/42454) ([Dan Roscigno](https://github.com/DanRoscigno)).
|
||||
* Remove code browser docs [#42455](https://github.com/ClickHouse/ClickHouse/pull/42455) ([Dan Roscigno](https://github.com/DanRoscigno)).
|
||||
* Better workaround for emitting .debug_aranges section [#42457](https://github.com/ClickHouse/ClickHouse/pull/42457) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Fix flaky test [#42459](https://github.com/ClickHouse/ClickHouse/pull/42459) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Fix UBSan report in Julian Day functions [#42464](https://github.com/ClickHouse/ClickHouse/pull/42464) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* rename filesystem_query_cache [#42472](https://github.com/ClickHouse/ClickHouse/pull/42472) ([Han Shukai](https://github.com/KinderRiven)).
|
||||
* Add convenience typedefs for Date/Date32/DateTime/DateTime64 columns [#42476](https://github.com/ClickHouse/ClickHouse/pull/42476) ([Robert Schulze](https://github.com/rschu1ze)).
|
||||
* Add error "Destination table is myself" to exception list in BC check [#42479](https://github.com/ClickHouse/ClickHouse/pull/42479) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Get current clickhouse version without sending query in BC check [#42483](https://github.com/ClickHouse/ClickHouse/pull/42483) ([Kruglov Pavel](https://github.com/Avogar)).
|
||||
* Fix logical error from welchTTest [#42487](https://github.com/ClickHouse/ClickHouse/pull/42487) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Attempt to fix abort from parallel parsing [#42496](https://github.com/ClickHouse/ClickHouse/pull/42496) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Increase threshold for using physical cores for `max_threads` [#42503](https://github.com/ClickHouse/ClickHouse/pull/42503) ([Nikita Taranov](https://github.com/nickitat)).
|
||||
* Add a test for [#16827](https://github.com/ClickHouse/ClickHouse/issues/16827) [#42511](https://github.com/ClickHouse/ClickHouse/pull/42511) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Add a test for [#13653](https://github.com/ClickHouse/ClickHouse/issues/13653) [#42512](https://github.com/ClickHouse/ClickHouse/pull/42512) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Fix aliases [#42514](https://github.com/ClickHouse/ClickHouse/pull/42514) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* tests: fix 00705_drop_create_merge_tree flakiness [#42522](https://github.com/ClickHouse/ClickHouse/pull/42522) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Fix sanitizer reports in integration tests [#42529](https://github.com/ClickHouse/ClickHouse/pull/42529) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Fix `KeeperTCPHandler` data race [#42532](https://github.com/ClickHouse/ClickHouse/pull/42532) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
* Disable `test_storage_nats`, because it's permanently broken [#42535](https://github.com/ClickHouse/ClickHouse/pull/42535) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Better logs in clickhouse-disks [#42549](https://github.com/ClickHouse/ClickHouse/pull/42549) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* add lib_fuzzer and lib_fuzzer_no_main to llvm-project build [#42550](https://github.com/ClickHouse/ClickHouse/pull/42550) ([Yakov Olkhovskiy](https://github.com/yakov-olkhovskiy)).
|
||||
* Some polishing: replicated merge tree [#42560](https://github.com/ClickHouse/ClickHouse/pull/42560) ([Igor Nikonov](https://github.com/devcrafter)).
|
||||
* Temporarily disable flaky `test_replicated_merge_tree_hdfs_zero_copy` [#42563](https://github.com/ClickHouse/ClickHouse/pull/42563) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Adapt internal data structures to 512-bit era [#42564](https://github.com/ClickHouse/ClickHouse/pull/42564) ([Nikita Taranov](https://github.com/nickitat)).
|
||||
* Fix strange code in date monotonicity [#42574](https://github.com/ClickHouse/ClickHouse/pull/42574) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Clear thread::id when ThreadFromGlobalPool exits. [#42577](https://github.com/ClickHouse/ClickHouse/pull/42577) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* ci/stress: fix memory limits overrides for hung check [#42585](https://github.com/ClickHouse/ClickHouse/pull/42585) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* tests: avoid model overlap for obfuscator [#42586](https://github.com/ClickHouse/ClickHouse/pull/42586) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Fix possible segfault in expression parser [#42598](https://github.com/ClickHouse/ClickHouse/pull/42598) ([Nikolay Degterinsky](https://github.com/evillique)).
|
||||
* Fix incorrect trace log line on dict reload [#42609](https://github.com/ClickHouse/ClickHouse/pull/42609) ([filimonov](https://github.com/filimonov)).
|
||||
* Fix flaky 02458_datediff_date32 test [#42611](https://github.com/ClickHouse/ClickHouse/pull/42611) ([Roman Vasin](https://github.com/rvasin)).
|
||||
* Revert revert 41268 disable s3 parallel write for part moves to disk s3 [#42617](https://github.com/ClickHouse/ClickHouse/pull/42617) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Try to fix data race on zookeeper vs DDLWorker at server shutdown. [#42620](https://github.com/ClickHouse/ClickHouse/pull/42620) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Add a template for installation issues [#42626](https://github.com/ClickHouse/ClickHouse/pull/42626) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Fix typo in cmake code related to fuzzing [#42627](https://github.com/ClickHouse/ClickHouse/pull/42627) ([Nikita Mikhaylov](https://github.com/nikitamikhaylov)).
|
||||
* Fix build [#42635](https://github.com/ClickHouse/ClickHouse/pull/42635) ([Anton Popov](https://github.com/CurtizJ)).
|
||||
* Add .rgignore for test data [#42639](https://github.com/ClickHouse/ClickHouse/pull/42639) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Fix flaky 02457_datediff_via_unix_epoch test [#42655](https://github.com/ClickHouse/ClickHouse/pull/42655) ([Roman Vasin](https://github.com/rvasin)).
|
||||
|
||||
29
docs/changelogs/v22.7.7.24-stable.md
Normal file
29
docs/changelogs/v22.7.7.24-stable.md
Normal file
@ -0,0 +1,29 @@
|
||||
---
|
||||
sidebar_position: 1
|
||||
sidebar_label: 2022
|
||||
---
|
||||
|
||||
# 2022 Changelog
|
||||
|
||||
### ClickHouse release v22.7.7.24-stable (02ad1f979a8) FIXME as compared to v22.7.6.74-stable (c00ffb3c11a)
|
||||
|
||||
#### Bug Fix
|
||||
* Backported in [#42433](https://github.com/ClickHouse/ClickHouse/issues/42433): - Choose correct aggregation method for LowCardinality with BigInt. [#42342](https://github.com/ClickHouse/ClickHouse/pull/42342) ([Duc Canh Le](https://github.com/canhld94)).
|
||||
|
||||
#### Build/Testing/Packaging Improvement
|
||||
* Backported in [#42329](https://github.com/ClickHouse/ClickHouse/issues/42329): Update cctz to the latest master, update tzdb to 2020e. [#42273](https://github.com/ClickHouse/ClickHouse/pull/42273) ([Dom Del Nano](https://github.com/ddelnano)).
|
||||
* Backported in [#42359](https://github.com/ClickHouse/ClickHouse/issues/42359): Update tzdata to 2022e to support the new timezone changes. Palestine transitions are now Saturdays at 02:00. Simplify three Ukraine zones into one. Jordan and Syria switch from +02/+03 with DST to year-round +03. (https://data.iana.org/time-zones/tzdb/NEWS). This closes [#42252](https://github.com/ClickHouse/ClickHouse/issues/42252). [#42327](https://github.com/ClickHouse/ClickHouse/pull/42327) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
|
||||
#### Bug Fix (user-visible misbehavior in official stable or prestable release)
|
||||
|
||||
* Backported in [#42268](https://github.com/ClickHouse/ClickHouse/issues/42268): Fix reusing of files > 4GB from base backup. [#42146](https://github.com/ClickHouse/ClickHouse/pull/42146) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Backported in [#42299](https://github.com/ClickHouse/ClickHouse/issues/42299): Fix a bug with projections and the `aggregate_functions_null_for_empty` setting. This bug is very rare and appears only if you enable the `aggregate_functions_null_for_empty` setting in the server's config. This closes [#41647](https://github.com/ClickHouse/ClickHouse/issues/41647). [#42198](https://github.com/ClickHouse/ClickHouse/pull/42198) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Backported in [#42386](https://github.com/ClickHouse/ClickHouse/issues/42386): `ALTER UPDATE` of attached part (with columns different from table schema) could create an invalid `columns.txt` metadata on disk. Reading from such part could fail with errors or return invalid data. Fixes [#42161](https://github.com/ClickHouse/ClickHouse/issues/42161). [#42319](https://github.com/ClickHouse/ClickHouse/pull/42319) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Backported in [#42498](https://github.com/ClickHouse/ClickHouse/issues/42498): Setting `additional_table_filters` were not applied to `Distributed` storage. Fixes [#41692](https://github.com/ClickHouse/ClickHouse/issues/41692). [#42322](https://github.com/ClickHouse/ClickHouse/pull/42322) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Backported in [#42593](https://github.com/ClickHouse/ClickHouse/issues/42593): This closes [#42453](https://github.com/ClickHouse/ClickHouse/issues/42453). [#42573](https://github.com/ClickHouse/ClickHouse/pull/42573) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
|
||||
#### NOT FOR CHANGELOG / INSIGNIFICANT
|
||||
|
||||
* Add a warning message to release.py script, require release type [#41975](https://github.com/ClickHouse/ClickHouse/pull/41975) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Revert [#27787](https://github.com/ClickHouse/ClickHouse/issues/27787) [#42136](https://github.com/ClickHouse/ClickHouse/pull/42136) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
|
||||
37
docs/changelogs/v22.8.7.34-lts.md
Normal file
37
docs/changelogs/v22.8.7.34-lts.md
Normal file
@ -0,0 +1,37 @@
|
||||
---
|
||||
sidebar_position: 1
|
||||
sidebar_label: 2022
|
||||
---
|
||||
|
||||
# 2022 Changelog
|
||||
|
||||
### ClickHouse release v22.8.7.34-lts (3c38e5e8ab9) FIXME as compared to v22.8.6.71-lts (7bf38a43e30)
|
||||
|
||||
#### Improvement
|
||||
* Backported in [#42096](https://github.com/ClickHouse/ClickHouse/issues/42096): Replace back `clickhouse su` command with `sudo -u` in start in order to respect limits in `/etc/security/limits.conf`. [#41847](https://github.com/ClickHouse/ClickHouse/pull/41847) ([Eugene Konkov](https://github.com/ekonkov)).
|
||||
|
||||
#### Bug Fix
|
||||
* Backported in [#42434](https://github.com/ClickHouse/ClickHouse/issues/42434): - Choose correct aggregation method for LowCardinality with BigInt. [#42342](https://github.com/ClickHouse/ClickHouse/pull/42342) ([Duc Canh Le](https://github.com/canhld94)).
|
||||
|
||||
#### Build/Testing/Packaging Improvement
|
||||
* Backported in [#42296](https://github.com/ClickHouse/ClickHouse/issues/42296): Update cctz to the latest master, update tzdb to 2020e. [#42273](https://github.com/ClickHouse/ClickHouse/pull/42273) ([Dom Del Nano](https://github.com/ddelnano)).
|
||||
* Backported in [#42360](https://github.com/ClickHouse/ClickHouse/issues/42360): Update tzdata to 2022e to support the new timezone changes. Palestine transitions are now Saturdays at 02:00. Simplify three Ukraine zones into one. Jordan and Syria switch from +02/+03 with DST to year-round +03. (https://data.iana.org/time-zones/tzdb/NEWS). This closes [#42252](https://github.com/ClickHouse/ClickHouse/issues/42252). [#42327](https://github.com/ClickHouse/ClickHouse/pull/42327) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
|
||||
#### Bug Fix (user-visible misbehavior in official stable or prestable release)
|
||||
|
||||
* Backported in [#42489](https://github.com/ClickHouse/ClickHouse/issues/42489): Removed skipping of mutations in unaffected partitions of `MergeTree` tables, because this feature never worked correctly and might cause resurrection of finished mutations. [#40589](https://github.com/ClickHouse/ClickHouse/pull/40589) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Backported in [#42121](https://github.com/ClickHouse/ClickHouse/issues/42121): Fixed "Part ... intersects part ..." error that might happen in extremely rare cases if replica was restarted just after detaching some part as broken. [#41741](https://github.com/ClickHouse/ClickHouse/pull/41741) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* - Prevent crash when passing wrong aggregation states to groupBitmap*. [#41972](https://github.com/ClickHouse/ClickHouse/pull/41972) ([Raúl Marín](https://github.com/Algunenano)).
|
||||
* - Fix read bytes/rows in X-ClickHouse-Summary with materialized views. [#41973](https://github.com/ClickHouse/ClickHouse/pull/41973) ([Raúl Marín](https://github.com/Algunenano)).
|
||||
* Backported in [#42269](https://github.com/ClickHouse/ClickHouse/issues/42269): Fix reusing of files > 4GB from base backup. [#42146](https://github.com/ClickHouse/ClickHouse/pull/42146) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Backported in [#42300](https://github.com/ClickHouse/ClickHouse/issues/42300): Fix a bug with projections and the `aggregate_functions_null_for_empty` setting. This bug is very rare and appears only if you enable the `aggregate_functions_null_for_empty` setting in the server's config. This closes [#41647](https://github.com/ClickHouse/ClickHouse/issues/41647). [#42198](https://github.com/ClickHouse/ClickHouse/pull/42198) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Backported in [#42387](https://github.com/ClickHouse/ClickHouse/issues/42387): `ALTER UPDATE` of attached part (with columns different from table schema) could create an invalid `columns.txt` metadata on disk. Reading from such part could fail with errors or return invalid data. Fixes [#42161](https://github.com/ClickHouse/ClickHouse/issues/42161). [#42319](https://github.com/ClickHouse/ClickHouse/pull/42319) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Backported in [#42499](https://github.com/ClickHouse/ClickHouse/issues/42499): Setting `additional_table_filters` were not applied to `Distributed` storage. Fixes [#41692](https://github.com/ClickHouse/ClickHouse/issues/41692). [#42322](https://github.com/ClickHouse/ClickHouse/pull/42322) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Backported in [#42571](https://github.com/ClickHouse/ClickHouse/issues/42571): Fix buffer overflow in the processing of Decimal data types. This closes [#42451](https://github.com/ClickHouse/ClickHouse/issues/42451). [#42465](https://github.com/ClickHouse/ClickHouse/pull/42465) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Backported in [#42594](https://github.com/ClickHouse/ClickHouse/issues/42594): This closes [#42453](https://github.com/ClickHouse/ClickHouse/issues/42453). [#42573](https://github.com/ClickHouse/ClickHouse/pull/42573) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
|
||||
#### NOT FOR CHANGELOG / INSIGNIFICANT
|
||||
|
||||
* Add a warning message to release.py script, require release type [#41975](https://github.com/ClickHouse/ClickHouse/pull/41975) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Revert [#27787](https://github.com/ClickHouse/ClickHouse/issues/27787) [#42136](https://github.com/ClickHouse/ClickHouse/pull/42136) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
|
||||
13
docs/changelogs/v22.8.8.3-lts.md
Normal file
13
docs/changelogs/v22.8.8.3-lts.md
Normal file
@ -0,0 +1,13 @@
|
||||
---
|
||||
sidebar_position: 1
|
||||
sidebar_label: 2022
|
||||
---
|
||||
|
||||
# 2022 Changelog
|
||||
|
||||
### ClickHouse release v22.8.8.3-lts (ac5a6cababc) FIXME as compared to v22.8.7.34-lts (3c38e5e8ab9)
|
||||
|
||||
#### Bug Fix (user-visible misbehavior in official stable or prestable release)
|
||||
|
||||
* Backported in [#42677](https://github.com/ClickHouse/ClickHouse/issues/42677): keeper-fix: fix race in accessing logs while snapshot is being installed. [#40627](https://github.com/ClickHouse/ClickHouse/pull/40627) ([Antonio Andelic](https://github.com/antonio2368)).
|
||||
|
||||
33
docs/changelogs/v22.9.4.32-stable.md
Normal file
33
docs/changelogs/v22.9.4.32-stable.md
Normal file
@ -0,0 +1,33 @@
|
||||
---
|
||||
sidebar_position: 1
|
||||
sidebar_label: 2022
|
||||
---
|
||||
|
||||
# 2022 Changelog
|
||||
|
||||
### ClickHouse release v22.9.4.32-stable (3db8bcf1a70) FIXME as compared to v22.9.3.18-stable (0cb4b15d2fa)
|
||||
|
||||
#### Bug Fix
|
||||
* Backported in [#42435](https://github.com/ClickHouse/ClickHouse/issues/42435): - Choose correct aggregation method for LowCardinality with BigInt. [#42342](https://github.com/ClickHouse/ClickHouse/pull/42342) ([Duc Canh Le](https://github.com/canhld94)).
|
||||
|
||||
#### Build/Testing/Packaging Improvement
|
||||
* Backported in [#42297](https://github.com/ClickHouse/ClickHouse/issues/42297): Update cctz to the latest master, update tzdb to 2020e. [#42273](https://github.com/ClickHouse/ClickHouse/pull/42273) ([Dom Del Nano](https://github.com/ddelnano)).
|
||||
* Backported in [#42361](https://github.com/ClickHouse/ClickHouse/issues/42361): Update tzdata to 2022e to support the new timezone changes. Palestine transitions are now Saturdays at 02:00. Simplify three Ukraine zones into one. Jordan and Syria switch from +02/+03 with DST to year-round +03. (https://data.iana.org/time-zones/tzdb/NEWS). This closes [#42252](https://github.com/ClickHouse/ClickHouse/issues/42252). [#42327](https://github.com/ClickHouse/ClickHouse/pull/42327) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
|
||||
#### Bug Fix (user-visible misbehavior in official stable or prestable release)
|
||||
|
||||
* Backported in [#42122](https://github.com/ClickHouse/ClickHouse/issues/42122): Fixed "Part ... intersects part ..." error that might happen in extremely rare cases if replica was restarted just after detaching some part as broken. [#41741](https://github.com/ClickHouse/ClickHouse/pull/41741) ([Alexander Tokmakov](https://github.com/tavplubix)).
|
||||
* Backported in [#41938](https://github.com/ClickHouse/ClickHouse/issues/41938): Don't allow to create or alter merge tree tables with virtual column name _row_exists, which is reserved for lightweight delete. Fixed [#41716](https://github.com/ClickHouse/ClickHouse/issues/41716). [#41763](https://github.com/ClickHouse/ClickHouse/pull/41763) ([Jianmei Zhang](https://github.com/zhangjmruc)).
|
||||
* Backported in [#42179](https://github.com/ClickHouse/ClickHouse/issues/42179): Fix reusing of files > 4GB from base backup. [#42146](https://github.com/ClickHouse/ClickHouse/pull/42146) ([Azat Khuzhin](https://github.com/azat)).
|
||||
* Backported in [#42301](https://github.com/ClickHouse/ClickHouse/issues/42301): Fix a bug with projections and the `aggregate_functions_null_for_empty` setting. This bug is very rare and appears only if you enable the `aggregate_functions_null_for_empty` setting in the server's config. This closes [#41647](https://github.com/ClickHouse/ClickHouse/issues/41647). [#42198](https://github.com/ClickHouse/ClickHouse/pull/42198) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Backported in [#42388](https://github.com/ClickHouse/ClickHouse/issues/42388): `ALTER UPDATE` of attached part (with columns different from table schema) could create an invalid `columns.txt` metadata on disk. Reading from such part could fail with errors or return invalid data. Fixes [#42161](https://github.com/ClickHouse/ClickHouse/issues/42161). [#42319](https://github.com/ClickHouse/ClickHouse/pull/42319) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Backported in [#42500](https://github.com/ClickHouse/ClickHouse/issues/42500): Setting `additional_table_filters` were not applied to `Distributed` storage. Fixes [#41692](https://github.com/ClickHouse/ClickHouse/issues/41692). [#42322](https://github.com/ClickHouse/ClickHouse/pull/42322) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
* Backported in [#42581](https://github.com/ClickHouse/ClickHouse/issues/42581): This reverts [#40217](https://github.com/ClickHouse/ClickHouse/issues/40217) which introduced a regression in date/time functions. [#42367](https://github.com/ClickHouse/ClickHouse/pull/42367) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Backported in [#42572](https://github.com/ClickHouse/ClickHouse/issues/42572): Fix buffer overflow in the processing of Decimal data types. This closes [#42451](https://github.com/ClickHouse/ClickHouse/issues/42451). [#42465](https://github.com/ClickHouse/ClickHouse/pull/42465) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
* Backported in [#42595](https://github.com/ClickHouse/ClickHouse/issues/42595): This closes [#42453](https://github.com/ClickHouse/ClickHouse/issues/42453). [#42573](https://github.com/ClickHouse/ClickHouse/pull/42573) ([Alexey Milovidov](https://github.com/alexey-milovidov)).
|
||||
|
||||
#### NOT FOR CHANGELOG / INSIGNIFICANT
|
||||
|
||||
* Add a warning message to release.py script, require release type [#41975](https://github.com/ClickHouse/ClickHouse/pull/41975) ([Mikhail f. Shiryaev](https://github.com/Felixoid)).
|
||||
* Revert [#27787](https://github.com/ClickHouse/ClickHouse/issues/27787) [#42136](https://github.com/ClickHouse/ClickHouse/pull/42136) ([Nikolai Kochetov](https://github.com/KochetovNicolai)).
|
||||
|
||||
@ -1,9 +1,12 @@
|
||||
---
|
||||
slug: /en/operations/backup
|
||||
sidebar_position: 49
|
||||
sidebar_label: Data backup and restore
|
||||
title: Data backup and restore
|
||||
---
|
||||
|
||||
[//]: # (This file is included in Manage > Backups)
|
||||
|
||||
- [Backup to a local disk](#backup-to-a-local-disk)
|
||||
- [Configuring backup/restore to use an S3 endpoint](#configuring-backuprestore-to-use-an-s3-endpoint)
|
||||
- [Backup/restore using an S3 disk](#backuprestore-using-an-s3-disk)
|
||||
- [Alternatives](#alternatives)
|
||||
|
||||
## Background
|
||||
|
||||
While [replication](../engines/table-engines/mergetree-family/replication.md) provides protection from hardware failures, it does not protect against human errors: accidental deletion of data, deletion of the wrong table or a table on the wrong cluster, and software bugs that result in incorrect data processing or data corruption. In many cases mistakes like these will affect all replicas. ClickHouse has built-in safeguards to prevent some types of mistakes — for example, by default [you can’t just drop tables with a MergeTree-like engine containing more than 50 Gb of data](server-configuration-parameters/settings.md#max-table-size-to-drop). However, these safeguards do not cover all possible cases and can be circumvented.
|
||||
|
||||
@ -15,7 +18,9 @@ Each company has different resources available and business requirements, so the
|
||||
Keep in mind that if you backed something up and never tried to restore it, chances are that restore will not work properly when you actually need it (or at least it will take longer than business can tolerate). So whatever backup approach you choose, make sure to automate the restore process as well, and practice it on a spare ClickHouse cluster regularly.
|
||||
:::
|
||||
|
||||
## Configure a backup destination
|
||||
## Backup to a local disk
|
||||
|
||||
### Configure a backup destination
|
||||
|
||||
In the examples below you will see the backup destination specified like `Disk('backups', '1.zip')`. To prepare the destination add a file to `/etc/clickhouse-server/config.d/backup_disk.xml` specifying the backup destination. For example, this file defines disk named `backups` and then adds that disk to the **backups > allowed_disk** list:
|
||||
|
||||
@ -39,7 +44,7 @@ In the examples below you will see the backup destination specified like `Disk('
|
||||
</clickhouse>
|
||||
```
|
||||
|
||||
## Parameters
|
||||
### Parameters
|
||||
|
||||
Backups can be either full or incremental, and can include tables (including materialized views, projections, and dictionaries), and databases. Backups can be synchronous (default) or asynchronous. They can be compressed. Backups can be password protected.
|
||||
|
||||
@ -52,7 +57,7 @@ The BACKUP and RESTORE statements take a list of DATABASE and TABLE names, a des
|
||||
- `password` for the file on disk
|
||||
- `base_backup`: the destination of the previous backup of this source. For example, `Disk('backups', '1.zip')`
|
||||
|
||||
## Usage examples
|
||||
### Usage examples
|
||||
|
||||
Backup and then restore a table:
|
||||
```
|
||||
@ -81,7 +86,7 @@ RESTORE TABLE test.table AS test.table2 FROM Disk('backups', '1.zip')
|
||||
BACKUP TABLE test.table3 AS test.table4 TO Disk('backups', '2.zip')
|
||||
```
|
||||
|
||||
## Incremental backups
|
||||
### Incremental backups
|
||||
|
||||
Incremental backups can be taken by specifying the `base_backup`.
|
||||
:::note
|
||||
@ -100,7 +105,7 @@ RESTORE TABLE test.table AS test.table2
|
||||
FROM Disk('backups', 'incremental-a.zip');
|
||||
```
|
||||
|
||||
## Assign a password to the backup
|
||||
### Assign a password to the backup
|
||||
|
||||
Backups written to disk can have a password applied to the file:
|
||||
```
|
||||
@ -116,7 +121,7 @@ RESTORE TABLE test.table
|
||||
SETTINGS password='qwerty'
|
||||
```
|
||||
|
||||
## Compression settings
|
||||
### Compression settings
|
||||
|
||||
If you would like to specify the compression method or level:
|
||||
```
|
||||
@ -125,14 +130,14 @@ BACKUP TABLE test.table
|
||||
SETTINGS compression_method='lzma', compression_level=3
|
||||
```
|
||||
|
||||
## Restore specific partitions
|
||||
### Restore specific partitions
|
||||
If specific partitions associated with a table need to be restored these can be specified. To restore partitions 1 and 4 from backup:
|
||||
```
|
||||
RESTORE TABLE test.table PARTITIONS '2', '3'
|
||||
FROM Disk('backups', 'filename.zip')
|
||||
```
|
||||
|
||||
## Check the status of backups
|
||||
### Check the status of backups
|
||||
|
||||
The backup command returns an `id` and `status`, and that `id` can be used to get the status of the backup. This is very useful to check the progress of long ASYNC backups. The example below shows a failure that happened when trying to overwrite an existing backup file:
|
||||
```sql
|
||||
@ -171,13 +176,118 @@ end_time: 2022-08-30 09:21:46
|
||||
1 row in set. Elapsed: 0.002 sec.
|
||||
```
|
||||
|
||||
## Backup to S3
|
||||
## Configuring BACKUP/RESTORE to use an S3 Endpoint
|
||||
|
||||
It is possible to `BACKUP`/`RESTORE` to S3, but this disk should be configured
|
||||
in a proper way, since by default you will need to backup metadata from local
|
||||
disk to make backup full.
|
||||
To write backups to an S3 bucket you need three pieces of information:
|
||||
- S3 endpoint,
|
||||
for example `https://mars-doc-test.s3.amazonaws.com/backup-S3/`
|
||||
- Access key ID,
|
||||
for example `ABC123`
|
||||
- Secret access key,
|
||||
for example `Abc+123`
|
||||
|
||||
First of all, you need to configure S3 disk in a special way:
|
||||
:::note
|
||||
Creating an S3 bucket is covered in [Use S3 Object Storage as a ClickHouse disk](/docs/en/integrations/data-ingestion/s3/configuring-s3-for-clickhouse-use.md), just come back to this doc after saving the policy, there is no need to configure ClickHouse to use the S3 bucket.
|
||||
:::
|
||||
|
||||
The destination for a backup will be specified like this:
|
||||
```
|
||||
S3('<S3 endpoint>/<directory>', '<Access key ID>', '<Secret access key>)
|
||||
```
|
||||
|
||||
```sql
|
||||
CREATE TABLE data
|
||||
(
|
||||
`key` Int,
|
||||
`value` String,
|
||||
`array` Array(String)
|
||||
)
|
||||
ENGINE = MergeTree
|
||||
ORDER BY tuple()
|
||||
```
|
||||
|
||||
```sql
|
||||
INSERT INTO data SELECT *
|
||||
FROM generateRandom('key Int, value String, array Array(String)')
|
||||
LIMIT 1000
|
||||
```
|
||||
|
||||
### Create a base (initial) backup
|
||||
|
||||
Incremental backups require a _base_ backup to start from, this example will be used
|
||||
later as the base backup. The first parameter of the S3 destination is the S3 endpoint followed by the directory within the bucket to use for this backup. In this example the directory is named `my_backup`.
|
||||
|
||||
```sql
|
||||
BACKUP TABLE data TO S3('https://mars-doc-test.s3.amazonaws.com/backup-S3/my_backup', 'ABC123', 'Abc+123')
|
||||
```
|
||||
|
||||
```response
|
||||
┌─id───────────────────────────────────┬─status─────────┐
|
||||
│ de442b75-a66c-4a3c-a193-f76f278c70f3 │ BACKUP_CREATED │
|
||||
└──────────────────────────────────────┴────────────────┘
|
||||
```
|
||||
|
||||
### Add more data
|
||||
|
||||
Incremental backups are populated with the difference between the base backup and the current content of the table being backed up. Add more data before taking the incremental backup:
|
||||
|
||||
```sql
|
||||
INSERT INTO data SELECT *
|
||||
FROM generateRandom('key Int, value String, array Array(String)')
|
||||
LIMIT 100
|
||||
```
|
||||
### Take an incremental backup
|
||||
|
||||
This backup command is similar to the base backup, but adds `SETTINGS base_backup` and the location of the base backup. Note that the destination for the incremental backup is not the same directory as the base, it is the same endpoint with a different target directory within the bucket. The base backup is in `my_backup`, and the incremental will be written to `my_incremental`:
|
||||
```sql
|
||||
BACKUP TABLE data TO S3('https://mars-doc-test.s3.amazonaws.com/backup-S3/my_incremental', 'ABC123', 'Abc+123') SETTINGS base_backup = S3('https://mars-doc-test.s3.amazonaws.com/backup-S3/my_backup', 'ABC123', 'Abc+123')
|
||||
```
|
||||
|
||||
```response
|
||||
┌─id───────────────────────────────────┬─status─────────┐
|
||||
│ f6cd3900-850f-41c9-94f1-0c4df33ea528 │ BACKUP_CREATED │
|
||||
└──────────────────────────────────────┴────────────────┘
|
||||
```
|
||||
### Restore from the incremental backup
|
||||
|
||||
This command restores the incremental backup into a new table, `data3`. Note that when an incremental backup is restored, the base backup is also included. Specify only the incremental backup when restoring:
|
||||
```sql
|
||||
RESTORE TABLE data AS data3 FROM S3('https://mars-doc-test.s3.amazonaws.com/backup-S3/my_incremental', 'ABC123', 'Abc+123')
|
||||
```
|
||||
|
||||
```response
|
||||
┌─id───────────────────────────────────┬─status───┐
|
||||
│ ff0c8c39-7dff-4324-a241-000796de11ca │ RESTORED │
|
||||
└──────────────────────────────────────┴──────────┘
|
||||
```
|
||||
|
||||
### Verify the count
|
||||
|
||||
There were two inserts into the original table `data`, one with 1,000 rows and one with 100 rows, for a total of 1,100. Verify that the restored table has 1,100 rows:
|
||||
```sql
|
||||
SELECT count()
|
||||
FROM data3
|
||||
```
|
||||
```response
|
||||
┌─count()─┐
|
||||
│ 1100 │
|
||||
└─────────┘
|
||||
```
|
||||
|
||||
### Verify the content
|
||||
This compares the content of the original table, `data` with the restored table `data3`:
|
||||
```sql
|
||||
SELECT throwIf((
|
||||
SELECT groupArray(tuple(*))
|
||||
FROM data
|
||||
) != (
|
||||
SELECT groupArray(tuple(*))
|
||||
FROM data3
|
||||
), 'Data does not match after BACKUP/RESTORE')
|
||||
```
|
||||
## BACKUP/RESTORE Using an S3 Disk
|
||||
|
||||
It is also possible to `BACKUP`/`RESTORE` to S3 by configuring an S3 disk in the ClickHouse storage configuration. Configure the disk like this by adding a file to `/etc/clickhouse-server/config.d`:
|
||||
|
||||
```xml
|
||||
<clickhouse>
|
||||
@ -194,7 +194,7 @@ To restore data from a backup, do the following:
|
||||
|
||||
Restoring from a backup does not require stopping the server.
|
||||
|
||||
For more information about backups and restoring data, see the [Data Backup](../../../operations/backup.md) section.
|
||||
For more information about backups and restoring data, see the [Data Backup](/docs/en/manage/backups.mdx) section.
|
||||
|
||||
## UNFREEZE PARTITION
|
||||
|
||||
|
||||
@ -87,14 +87,15 @@ SETTINGS
|
||||
|
||||
<summary>Устаревший способ создания таблицы</summary>
|
||||
|
||||
:::note "Attention"
|
||||
Не используйте этот метод в новых проектах. По возможности переключите старые проекты на метод, описанный выше.
|
||||
:::note "Attention"
|
||||
Не используйте этот метод в новых проектах. По возможности переключите старые проекты на метод, описанный выше.
|
||||
:::
|
||||
|
||||
``` sql
|
||||
Kafka(kafka_broker_list, kafka_topic_list, kafka_group_name, kafka_format
|
||||
[, kafka_row_delimiter, kafka_schema, kafka_num_consumers, kafka_skip_broken_messages])
|
||||
```
|
||||
:::
|
||||
|
||||
</details>
|
||||
|
||||
## Описание {#opisanie}
|
||||
|
||||
@ -39,9 +39,10 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
|
||||
<summary>Устаревший способ создания таблицы</summary>
|
||||
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ описанный выше.
|
||||
:::
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ описанный выше.
|
||||
:::
|
||||
|
||||
``` sql
|
||||
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
(
|
||||
|
||||
@ -43,9 +43,10 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
|
||||
<summary>Устаревший способ создания таблицы</summary>
|
||||
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ описанный выше.
|
||||
:::
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ описанный выше.
|
||||
:::
|
||||
|
||||
``` sql
|
||||
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
(
|
||||
@ -59,7 +60,7 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
|
||||
- `sign` — Имя столбца с типом строки: `1` — строка состояния, `-1` — строка отмены состояния.
|
||||
|
||||
Тип данных столбца — `Int8`.
|
||||
Тип данных столбца — `Int8`.
|
||||
|
||||
</details>
|
||||
|
||||
|
||||
@ -55,9 +55,10 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
|
||||
<summary>Устаревший способ создания таблицы</summary>
|
||||
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ описанный выше.
|
||||
:::
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ описанный выше.
|
||||
:::
|
||||
|
||||
``` sql
|
||||
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
(
|
||||
|
||||
@ -115,9 +115,10 @@ ENGINE MergeTree() PARTITION BY toYYYYMM(EventDate) ORDER BY (CounterID, EventDa
|
||||
|
||||
<summary>Устаревший способ создания таблицы</summary>
|
||||
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ, описанный выше.
|
||||
:::
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ, описанный выше.
|
||||
:::
|
||||
|
||||
``` sql
|
||||
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
(
|
||||
|
||||
@ -42,9 +42,10 @@ CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
|
||||
<summary>Устаревший способ создания таблицы</summary>
|
||||
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ описанный выше.
|
||||
:::
|
||||
:::note "Attention"
|
||||
Не используйте этот способ в новых проектах и по возможности переведите старые проекты на способ описанный выше.
|
||||
:::
|
||||
|
||||
``` sql
|
||||
CREATE TABLE [IF NOT EXISTS] [db.]table_name [ON CLUSTER cluster]
|
||||
(
|
||||
|
||||
@ -316,9 +316,9 @@ SELECT toStartOfISOYear(toDate('2017-01-01')) AS ISOYear20170101;
|
||||
Возвращается дата.
|
||||
|
||||
:::note "Attention"
|
||||
Возвращаемое значение для некорректных дат зависит от реализации. ClickHouse может вернуть нулевую дату, выбросить исключение, или выполнить «естественное» перетекание дат между месяцами.
|
||||
Возвращаемое значение для некорректных дат зависит от реализации. ClickHouse может вернуть нулевую дату, выбросить исключение, или выполнить «естественное» перетекание дат между месяцами.
|
||||
:::
|
||||
|
||||
|
||||
## toMonday {#tomonday}
|
||||
|
||||
Округляет дату или дату-с-временем вниз до ближайшего понедельника.
|
||||
|
||||
@ -122,9 +122,9 @@ FROM t_null
|
||||
|
||||
Существует два варианта IN-ов с подзапросами (аналогично для JOIN-ов): обычный `IN` / `JOIN` и `GLOBAL IN` / `GLOBAL JOIN`. Они отличаются способом выполнения при распределённой обработке запроса.
|
||||
|
||||
:::note "Attention"
|
||||
Помните, что алгоритмы, описанные ниже, могут работать иначе в зависимости от [настройки](../../operations/settings/settings.md) `distributed_product_mode`.
|
||||
:::
|
||||
:::note "Attention"
|
||||
Помните, что алгоритмы, описанные ниже, могут работать иначе в зависимости от [настройки](../../operations/settings/settings.md) `distributed_product_mode`.
|
||||
:::
|
||||
При использовании обычного IN-а, запрос отправляется на удалённые серверы, и на каждом из них выполняются подзапросы в секциях `IN` / `JOIN`.
|
||||
|
||||
При использовании `GLOBAL IN` / `GLOBAL JOIN-а`, сначала выполняются все подзапросы для `GLOBAL IN` / `GLOBAL JOIN-ов`, и результаты складываются во временные таблицы. Затем эти временные таблицы передаются на каждый удалённый сервер, и на них выполняются запросы, с использованием этих переданных временных данных.
|
||||
|
||||
@ -1,6 +1,10 @@
|
||||
#pragma once
|
||||
|
||||
#include <Interpreters/Cluster.h>
|
||||
#include <base/types.h>
|
||||
|
||||
#include <Poco/Util/AbstractConfiguration.h>
|
||||
|
||||
#include <utility>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
@ -8,21 +12,4 @@ namespace DB
|
||||
|
||||
using DatabaseAndTableName = std::pair<String, String>;
|
||||
using ListOfDatabasesAndTableNames = std::vector<DatabaseAndTableName>;
|
||||
|
||||
/// Hierarchical description of the tasks
|
||||
struct ShardPartitionPiece;
|
||||
struct ShardPartition;
|
||||
struct TaskShard;
|
||||
struct TaskTable;
|
||||
struct TaskCluster;
|
||||
struct ClusterPartition;
|
||||
|
||||
using PartitionPieces = std::vector<ShardPartitionPiece>;
|
||||
using TasksPartition = std::map<String, ShardPartition, std::greater<>>;
|
||||
using ShardInfo = Cluster::ShardInfo;
|
||||
using TaskShardPtr = std::shared_ptr<TaskShard>;
|
||||
using TasksShard = std::vector<TaskShardPtr>;
|
||||
using TasksTable = std::list<TaskTable>;
|
||||
using ClusterPartitions = std::map<String, ClusterPartition, std::greater<>>;
|
||||
}
|
||||
|
||||
|
||||
@ -1,7 +1,13 @@
|
||||
set(CLICKHOUSE_COPIER_SOURCES
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/ClusterCopierApp.cpp"
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/ClusterCopier.cpp"
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/Internals.cpp")
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/Internals.cpp"
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/ShardPartition.cpp"
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/ShardPartitionPiece.cpp"
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/StatusAccumulator.cpp"
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/TaskCluster.cpp"
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/TaskShard.cpp"
|
||||
"${CMAKE_CURRENT_SOURCE_DIR}/TaskTable.cpp")
|
||||
|
||||
set (CLICKHOUSE_COPIER_LINK
|
||||
PRIVATE
|
||||
|
||||
@ -3,7 +3,8 @@
|
||||
#include "Aliases.h"
|
||||
#include "Internals.h"
|
||||
#include "TaskCluster.h"
|
||||
#include "TaskTableAndShard.h"
|
||||
#include "TaskShard.h"
|
||||
#include "TaskTable.h"
|
||||
#include "ShardPartition.h"
|
||||
#include "ShardPartitionPiece.h"
|
||||
#include "ZooKeeperStaff.h"
|
||||
|
||||
@ -1,17 +1,22 @@
|
||||
#pragma once
|
||||
|
||||
#include "Aliases.h"
|
||||
#include <base/types.h>
|
||||
#include <map>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
/// Contains info about all shards that contain a partition
|
||||
struct ClusterPartition
|
||||
{
|
||||
double elapsed_time_seconds = 0;
|
||||
UInt64 bytes_copied = 0;
|
||||
UInt64 rows_copied = 0;
|
||||
UInt64 blocks_copied = 0;
|
||||
|
||||
UInt64 total_tries = 0;
|
||||
};
|
||||
/// Contains info about all shards that contain a partition
|
||||
struct ClusterPartition
|
||||
{
|
||||
double elapsed_time_seconds = 0;
|
||||
UInt64 bytes_copied = 0;
|
||||
UInt64 rows_copied = 0;
|
||||
UInt64 blocks_copied = 0;
|
||||
|
||||
UInt64 total_tries = 0;
|
||||
};
|
||||
|
||||
using ClusterPartitions = std::map<String, ClusterPartition, std::greater<>>;
|
||||
|
||||
}
|
||||
|
||||
70
programs/copier/ShardPartition.cpp
Normal file
70
programs/copier/ShardPartition.cpp
Normal file
@ -0,0 +1,70 @@
|
||||
#include "ShardPartition.h"
|
||||
|
||||
#include "TaskShard.h"
|
||||
#include "TaskTable.h"
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
ShardPartition::ShardPartition(TaskShard & parent, String name_quoted_, size_t number_of_splits)
|
||||
: task_shard(parent)
|
||||
, name(std::move(name_quoted_))
|
||||
{
|
||||
pieces.reserve(number_of_splits);
|
||||
}
|
||||
|
||||
String ShardPartition::getPartitionCleanStartPath() const
|
||||
{
|
||||
return getPartitionPath() + "/clean_start";
|
||||
}
|
||||
|
||||
String ShardPartition::getPartitionPieceCleanStartPath(size_t current_piece_number) const
|
||||
{
|
||||
assert(current_piece_number < task_shard.task_table.number_of_splits);
|
||||
return getPartitionPiecePath(current_piece_number) + "/clean_start";
|
||||
}
|
||||
|
||||
String ShardPartition::getPartitionPath() const
|
||||
{
|
||||
return task_shard.task_table.getPartitionPath(name);
|
||||
}
|
||||
|
||||
String ShardPartition::getPartitionPiecePath(size_t current_piece_number) const
|
||||
{
|
||||
assert(current_piece_number < task_shard.task_table.number_of_splits);
|
||||
return task_shard.task_table.getPartitionPiecePath(name, current_piece_number);
|
||||
}
|
||||
|
||||
String ShardPartition::getShardStatusPath() const
|
||||
{
|
||||
// schema: /<root...>/tables/<table>/<partition>/shards/<shard>
|
||||
// e.g. /root/table_test.hits/201701/shards/1
|
||||
return getPartitionShardsPath() + "/" + toString(task_shard.numberInCluster());
|
||||
}
|
||||
|
||||
String ShardPartition::getPartitionShardsPath() const
|
||||
{
|
||||
return getPartitionPath() + "/shards";
|
||||
}
|
||||
|
||||
String ShardPartition::getPartitionActiveWorkersPath() const
|
||||
{
|
||||
return getPartitionPath() + "/partition_active_workers";
|
||||
}
|
||||
|
||||
String ShardPartition::getActiveWorkerPath() const
|
||||
{
|
||||
return getPartitionActiveWorkersPath() + "/" + toString(task_shard.numberInCluster());
|
||||
}
|
||||
|
||||
String ShardPartition::getCommonPartitionIsDirtyPath() const
|
||||
{
|
||||
return getPartitionPath() + "/is_dirty";
|
||||
}
|
||||
|
||||
String ShardPartition::getCommonPartitionIsCleanedPath() const
|
||||
{
|
||||
return getCommonPartitionIsDirtyPath() + "/cleaned";
|
||||
}
|
||||
|
||||
}
|
||||
@ -1,19 +1,23 @@
|
||||
#pragma once
|
||||
|
||||
#include "Aliases.h"
|
||||
#include "TaskTableAndShard.h"
|
||||
#include "ShardPartitionPiece.h"
|
||||
|
||||
#include <base/types.h>
|
||||
|
||||
#include <map>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
struct TaskShard;
|
||||
|
||||
/// Just destination partition of a shard
|
||||
/// I don't know what this comment means.
|
||||
/// In short, when we discovered what shards contain currently processing partition,
|
||||
/// This class describes a partition (name) that is stored on the shard (parent).
|
||||
struct ShardPartition
|
||||
{
|
||||
ShardPartition(TaskShard &parent, String name_quoted_, size_t number_of_splits = 10)
|
||||
: task_shard(parent), name(std::move(name_quoted_)) { pieces.reserve(number_of_splits); }
|
||||
ShardPartition(TaskShard &parent, String name_quoted_, size_t number_of_splits = 10);
|
||||
|
||||
String getPartitionPath() const;
|
||||
|
||||
@ -45,58 +49,6 @@ struct ShardPartition
|
||||
String name;
|
||||
};
|
||||
|
||||
inline String ShardPartition::getPartitionCleanStartPath() const
|
||||
{
|
||||
return getPartitionPath() + "/clean_start";
|
||||
}
|
||||
|
||||
inline String ShardPartition::getPartitionPieceCleanStartPath(size_t current_piece_number) const
|
||||
{
|
||||
assert(current_piece_number < task_shard.task_table.number_of_splits);
|
||||
return getPartitionPiecePath(current_piece_number) + "/clean_start";
|
||||
}
|
||||
|
||||
inline String ShardPartition::getPartitionPath() const
|
||||
{
|
||||
return task_shard.task_table.getPartitionPath(name);
|
||||
}
|
||||
|
||||
inline String ShardPartition::getPartitionPiecePath(size_t current_piece_number) const
|
||||
{
|
||||
assert(current_piece_number < task_shard.task_table.number_of_splits);
|
||||
return task_shard.task_table.getPartitionPiecePath(name, current_piece_number);
|
||||
}
|
||||
|
||||
inline String ShardPartition::getShardStatusPath() const
|
||||
{
|
||||
// schema: /<root...>/tables/<table>/<partition>/shards/<shard>
|
||||
// e.g. /root/table_test.hits/201701/shards/1
|
||||
return getPartitionShardsPath() + "/" + toString(task_shard.numberInCluster());
|
||||
}
|
||||
|
||||
inline String ShardPartition::getPartitionShardsPath() const
|
||||
{
|
||||
return getPartitionPath() + "/shards";
|
||||
}
|
||||
|
||||
inline String ShardPartition::getPartitionActiveWorkersPath() const
|
||||
{
|
||||
return getPartitionPath() + "/partition_active_workers";
|
||||
}
|
||||
|
||||
inline String ShardPartition::getActiveWorkerPath() const
|
||||
{
|
||||
return getPartitionActiveWorkersPath() + "/" + toString(task_shard.numberInCluster());
|
||||
}
|
||||
|
||||
inline String ShardPartition::getCommonPartitionIsDirtyPath() const
|
||||
{
|
||||
return getPartitionPath() + "/is_dirty";
|
||||
}
|
||||
|
||||
inline String ShardPartition::getCommonPartitionIsCleanedPath() const
|
||||
{
|
||||
return getCommonPartitionIsDirtyPath() + "/cleaned";
|
||||
}
|
||||
using TasksPartition = std::map<String, ShardPartition, std::greater<>>;
|
||||
|
||||
}
|
||||
|
||||
64
programs/copier/ShardPartitionPiece.cpp
Normal file
64
programs/copier/ShardPartitionPiece.cpp
Normal file
@ -0,0 +1,64 @@
|
||||
#include "ShardPartitionPiece.h"
|
||||
|
||||
#include "ShardPartition.h"
|
||||
#include "TaskShard.h"
|
||||
|
||||
#include <IO/WriteHelpers.h>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
ShardPartitionPiece::ShardPartitionPiece(ShardPartition & parent, size_t current_piece_number_, bool is_present_piece_)
|
||||
: is_absent_piece(!is_present_piece_)
|
||||
, current_piece_number(current_piece_number_)
|
||||
, shard_partition(parent)
|
||||
{
|
||||
}
|
||||
|
||||
String ShardPartitionPiece::getPartitionPiecePath() const
|
||||
{
|
||||
return shard_partition.getPartitionPath() + "/piece_" + toString(current_piece_number);
|
||||
}
|
||||
|
||||
String ShardPartitionPiece::getPartitionPieceCleanStartPath() const
|
||||
{
|
||||
return getPartitionPiecePath() + "/clean_start";
|
||||
}
|
||||
|
||||
String ShardPartitionPiece::getPartitionPieceIsDirtyPath() const
|
||||
{
|
||||
return getPartitionPiecePath() + "/is_dirty";
|
||||
}
|
||||
|
||||
String ShardPartitionPiece::getPartitionPieceIsCleanedPath() const
|
||||
{
|
||||
return getPartitionPieceIsDirtyPath() + "/cleaned";
|
||||
}
|
||||
|
||||
String ShardPartitionPiece::getPartitionPieceActiveWorkersPath() const
|
||||
{
|
||||
return getPartitionPiecePath() + "/partition_piece_active_workers";
|
||||
}
|
||||
|
||||
String ShardPartitionPiece::getActiveWorkerPath() const
|
||||
{
|
||||
return getPartitionPieceActiveWorkersPath() + "/" + toString(shard_partition.task_shard.numberInCluster());
|
||||
}
|
||||
|
||||
/// On what shards do we have current partition.
|
||||
String ShardPartitionPiece::getPartitionPieceShardsPath() const
|
||||
{
|
||||
return getPartitionPiecePath() + "/shards";
|
||||
}
|
||||
|
||||
String ShardPartitionPiece::getShardStatusPath() const
|
||||
{
|
||||
return getPartitionPieceShardsPath() + "/" + toString(shard_partition.task_shard.numberInCluster());
|
||||
}
|
||||
|
||||
String ShardPartitionPiece::getPartitionPieceCleanerPath() const
|
||||
{
|
||||
return getPartitionPieceIsDirtyPath() + "/cleaner";
|
||||
}
|
||||
|
||||
}
|
||||
@ -1,16 +1,15 @@
|
||||
#pragma once
|
||||
|
||||
#include "Internals.h"
|
||||
#include <base/types.h>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
struct ShardPartition;
|
||||
|
||||
struct ShardPartitionPiece
|
||||
{
|
||||
|
||||
ShardPartitionPiece(ShardPartition &parent, size_t current_piece_number_, bool is_present_piece_)
|
||||
: is_absent_piece(!is_present_piece_), current_piece_number(current_piece_number_),
|
||||
shard_partition(parent) {}
|
||||
ShardPartitionPiece(ShardPartition & parent, size_t current_piece_number_, bool is_present_piece_);
|
||||
|
||||
String getPartitionPiecePath() const;
|
||||
|
||||
@ -37,52 +36,6 @@ struct ShardPartitionPiece
|
||||
ShardPartition & shard_partition;
|
||||
};
|
||||
|
||||
|
||||
inline String ShardPartitionPiece::getPartitionPiecePath() const
|
||||
{
|
||||
return shard_partition.getPartitionPath() + "/piece_" + toString(current_piece_number);
|
||||
}
|
||||
|
||||
inline String ShardPartitionPiece::getPartitionPieceCleanStartPath() const
|
||||
{
|
||||
return getPartitionPiecePath() + "/clean_start";
|
||||
}
|
||||
|
||||
inline String ShardPartitionPiece::getPartitionPieceIsDirtyPath() const
|
||||
{
|
||||
return getPartitionPiecePath() + "/is_dirty";
|
||||
}
|
||||
|
||||
inline String ShardPartitionPiece::getPartitionPieceIsCleanedPath() const
|
||||
{
|
||||
return getPartitionPieceIsDirtyPath() + "/cleaned";
|
||||
}
|
||||
|
||||
inline String ShardPartitionPiece::getPartitionPieceActiveWorkersPath() const
|
||||
{
|
||||
return getPartitionPiecePath() + "/partition_piece_active_workers";
|
||||
}
|
||||
|
||||
inline String ShardPartitionPiece::getActiveWorkerPath() const
|
||||
{
|
||||
return getPartitionPieceActiveWorkersPath() + "/" + toString(shard_partition.task_shard.numberInCluster());
|
||||
}
|
||||
|
||||
/// On what shards do we have current partition.
|
||||
inline String ShardPartitionPiece::getPartitionPieceShardsPath() const
|
||||
{
|
||||
return getPartitionPiecePath() + "/shards";
|
||||
}
|
||||
|
||||
inline String ShardPartitionPiece::getShardStatusPath() const
|
||||
{
|
||||
return getPartitionPieceShardsPath() + "/" + toString(shard_partition.task_shard.numberInCluster());
|
||||
}
|
||||
|
||||
inline String ShardPartitionPiece::getPartitionPieceCleanerPath() const
|
||||
{
|
||||
return getPartitionPieceIsDirtyPath() + "/cleaner";
|
||||
}
|
||||
|
||||
using PartitionPieces = std::vector<ShardPartitionPiece>;
|
||||
|
||||
}
|
||||
|
||||
48
programs/copier/StatusAccumulator.cpp
Normal file
48
programs/copier/StatusAccumulator.cpp
Normal file
@ -0,0 +1,48 @@
|
||||
#include "StatusAccumulator.h"
|
||||
|
||||
#include <Poco/JSON/Parser.h>
|
||||
#include <Poco/JSON/JSON.h>
|
||||
#include <Poco/JSON/Object.h>
|
||||
#include <Poco/JSON/Stringifier.h>
|
||||
|
||||
#include <iostream>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
StatusAccumulator::MapPtr StatusAccumulator::fromJSON(String state_json)
|
||||
{
|
||||
Poco::JSON::Parser parser;
|
||||
auto state = parser.parse(state_json).extract<Poco::JSON::Object::Ptr>();
|
||||
MapPtr result_ptr = std::make_shared<Map>();
|
||||
for (const auto & table_name : state->getNames())
|
||||
{
|
||||
auto table_status_json = state->getValue<String>(table_name);
|
||||
auto table_status = parser.parse(table_status_json).extract<Poco::JSON::Object::Ptr>();
|
||||
/// Map entry will be created if it is absent
|
||||
auto & map_table_status = (*result_ptr)[table_name];
|
||||
map_table_status.all_partitions_count += table_status->getValue<size_t>("all_partitions_count");
|
||||
map_table_status.processed_partitions_count += table_status->getValue<size_t>("processed_partitions_count");
|
||||
}
|
||||
return result_ptr;
|
||||
}
|
||||
|
||||
String StatusAccumulator::serializeToJSON(MapPtr statuses)
|
||||
{
|
||||
Poco::JSON::Object result_json;
|
||||
for (const auto & [table_name, table_status] : *statuses)
|
||||
{
|
||||
Poco::JSON::Object status_json;
|
||||
status_json.set("all_partitions_count", table_status.all_partitions_count);
|
||||
status_json.set("processed_partitions_count", table_status.processed_partitions_count);
|
||||
|
||||
result_json.set(table_name, status_json);
|
||||
}
|
||||
std::ostringstream oss; // STYLE_CHECK_ALLOW_STD_STRING_STREAM
|
||||
oss.exceptions(std::ios::failbit);
|
||||
Poco::JSON::Stringifier::stringify(result_json, oss);
|
||||
auto result = oss.str();
|
||||
return result;
|
||||
}
|
||||
|
||||
}
|
||||
@ -1,65 +1,27 @@
|
||||
#pragma once
|
||||
|
||||
#include <base/types.h>
|
||||
|
||||
#include <Poco/JSON/Parser.h>
|
||||
#include <Poco/JSON/JSON.h>
|
||||
#include <Poco/JSON/Object.h>
|
||||
#include <Poco/JSON/Stringifier.h>
|
||||
|
||||
#include <unordered_map>
|
||||
#include <memory>
|
||||
#include <string>
|
||||
#include <iostream>
|
||||
#include <unordered_map>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
class StatusAccumulator
|
||||
{
|
||||
public:
|
||||
struct TableStatus
|
||||
{
|
||||
size_t all_partitions_count;
|
||||
size_t processed_partitions_count;
|
||||
};
|
||||
public:
|
||||
struct TableStatus
|
||||
{
|
||||
size_t all_partitions_count;
|
||||
size_t processed_partitions_count;
|
||||
};
|
||||
|
||||
using Map = std::unordered_map<std::string, TableStatus>;
|
||||
using MapPtr = std::shared_ptr<Map>;
|
||||
using Map = std::unordered_map<String, TableStatus>;
|
||||
using MapPtr = std::shared_ptr<Map>;
|
||||
|
||||
static MapPtr fromJSON(std::string state_json)
|
||||
{
|
||||
Poco::JSON::Parser parser;
|
||||
auto state = parser.parse(state_json).extract<Poco::JSON::Object::Ptr>();
|
||||
MapPtr result_ptr = std::make_shared<Map>();
|
||||
for (const auto & table_name : state->getNames())
|
||||
{
|
||||
auto table_status_json = state->getValue<std::string>(table_name);
|
||||
auto table_status = parser.parse(table_status_json).extract<Poco::JSON::Object::Ptr>();
|
||||
/// Map entry will be created if it is absent
|
||||
auto & map_table_status = (*result_ptr)[table_name];
|
||||
map_table_status.all_partitions_count += table_status->getValue<size_t>("all_partitions_count");
|
||||
map_table_status.processed_partitions_count += table_status->getValue<size_t>("processed_partitions_count");
|
||||
}
|
||||
return result_ptr;
|
||||
}
|
||||
|
||||
static std::string serializeToJSON(MapPtr statuses)
|
||||
{
|
||||
Poco::JSON::Object result_json;
|
||||
for (const auto & [table_name, table_status] : *statuses)
|
||||
{
|
||||
Poco::JSON::Object status_json;
|
||||
status_json.set("all_partitions_count", table_status.all_partitions_count);
|
||||
status_json.set("processed_partitions_count", table_status.processed_partitions_count);
|
||||
|
||||
result_json.set(table_name, status_json);
|
||||
}
|
||||
std::ostringstream oss; // STYLE_CHECK_ALLOW_STD_STRING_STREAM
|
||||
oss.exceptions(std::ios::failbit);
|
||||
Poco::JSON::Stringifier::stringify(result_json, oss);
|
||||
auto result = oss.str();
|
||||
return result;
|
||||
}
|
||||
static MapPtr fromJSON(String state_json);
|
||||
static String serializeToJSON(MapPtr statuses);
|
||||
};
|
||||
|
||||
}
|
||||
|
||||
74
programs/copier/TaskCluster.cpp
Normal file
74
programs/copier/TaskCluster.cpp
Normal file
@ -0,0 +1,74 @@
|
||||
#include "TaskCluster.h"
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
namespace ErrorCodes
|
||||
{
|
||||
extern const int BAD_ARGUMENTS;
|
||||
}
|
||||
|
||||
TaskCluster::TaskCluster(const String & task_zookeeper_path_, const String & default_local_database_)
|
||||
: task_zookeeper_path(task_zookeeper_path_)
|
||||
, default_local_database(default_local_database_)
|
||||
{}
|
||||
|
||||
void DB::TaskCluster::loadTasks(const Poco::Util::AbstractConfiguration & config, const String & base_key)
|
||||
{
|
||||
String prefix = base_key.empty() ? "" : base_key + ".";
|
||||
|
||||
clusters_prefix = prefix + "remote_servers";
|
||||
if (!config.has(clusters_prefix))
|
||||
throw Exception("You should specify list of clusters in " + clusters_prefix, ErrorCodes::BAD_ARGUMENTS);
|
||||
|
||||
Poco::Util::AbstractConfiguration::Keys tables_keys;
|
||||
config.keys(prefix + "tables", tables_keys);
|
||||
|
||||
for (const auto & table_key : tables_keys)
|
||||
{
|
||||
table_tasks.emplace_back(*this, config, prefix + "tables", table_key);
|
||||
}
|
||||
}
|
||||
|
||||
void DB::TaskCluster::reloadSettings(const Poco::Util::AbstractConfiguration & config, const String & base_key)
|
||||
{
|
||||
String prefix = base_key.empty() ? "" : base_key + ".";
|
||||
|
||||
max_workers = config.getUInt64(prefix + "max_workers");
|
||||
|
||||
settings_common = Settings();
|
||||
if (config.has(prefix + "settings"))
|
||||
settings_common.loadSettingsFromConfig(prefix + "settings", config);
|
||||
|
||||
settings_common.prefer_localhost_replica = false;
|
||||
|
||||
settings_pull = settings_common;
|
||||
if (config.has(prefix + "settings_pull"))
|
||||
settings_pull.loadSettingsFromConfig(prefix + "settings_pull", config);
|
||||
|
||||
settings_push = settings_common;
|
||||
if (config.has(prefix + "settings_push"))
|
||||
settings_push.loadSettingsFromConfig(prefix + "settings_push", config);
|
||||
|
||||
auto set_default_value = [] (auto && setting, auto && default_value)
|
||||
{
|
||||
setting = setting.changed ? setting.value : default_value;
|
||||
};
|
||||
|
||||
/// Override important settings
|
||||
settings_pull.readonly = 1;
|
||||
settings_pull.prefer_localhost_replica = false;
|
||||
settings_push.insert_distributed_sync = true;
|
||||
settings_push.prefer_localhost_replica = false;
|
||||
|
||||
set_default_value(settings_pull.load_balancing, LoadBalancing::NEAREST_HOSTNAME);
|
||||
set_default_value(settings_pull.max_threads, 1);
|
||||
set_default_value(settings_pull.max_block_size, 8192UL);
|
||||
set_default_value(settings_pull.preferred_block_size_bytes, 0);
|
||||
|
||||
set_default_value(settings_push.insert_distributed_timeout, 0);
|
||||
set_default_value(settings_push.replication_alter_partitions_sync, 2);
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -1,21 +1,20 @@
|
||||
#pragma once
|
||||
|
||||
#include "Aliases.h"
|
||||
#include "TaskTable.h"
|
||||
|
||||
#include <Core/Settings.h>
|
||||
#include <base/types.h>
|
||||
|
||||
#include <Poco/Util/AbstractConfiguration.h>
|
||||
|
||||
#include <random>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
namespace ErrorCodes
|
||||
{
|
||||
extern const int BAD_ARGUMENTS;
|
||||
}
|
||||
|
||||
struct TaskCluster
|
||||
{
|
||||
TaskCluster(const String & task_zookeeper_path_, const String & default_local_database_)
|
||||
: task_zookeeper_path(task_zookeeper_path_)
|
||||
, default_local_database(default_local_database_)
|
||||
{}
|
||||
TaskCluster(const String & task_zookeeper_path_, const String & default_local_database_);
|
||||
|
||||
void loadTasks(const Poco::Util::AbstractConfiguration & config, const String & base_key = "");
|
||||
|
||||
@ -50,61 +49,4 @@ struct TaskCluster
|
||||
pcg64 random_engine;
|
||||
};
|
||||
|
||||
inline void DB::TaskCluster::loadTasks(const Poco::Util::AbstractConfiguration & config, const String & base_key)
|
||||
{
|
||||
String prefix = base_key.empty() ? "" : base_key + ".";
|
||||
|
||||
clusters_prefix = prefix + "remote_servers";
|
||||
if (!config.has(clusters_prefix))
|
||||
throw Exception("You should specify list of clusters in " + clusters_prefix, ErrorCodes::BAD_ARGUMENTS);
|
||||
|
||||
Poco::Util::AbstractConfiguration::Keys tables_keys;
|
||||
config.keys(prefix + "tables", tables_keys);
|
||||
|
||||
for (const auto & table_key : tables_keys)
|
||||
{
|
||||
table_tasks.emplace_back(*this, config, prefix + "tables", table_key);
|
||||
}
|
||||
}
|
||||
|
||||
inline void DB::TaskCluster::reloadSettings(const Poco::Util::AbstractConfiguration & config, const String & base_key)
|
||||
{
|
||||
String prefix = base_key.empty() ? "" : base_key + ".";
|
||||
|
||||
max_workers = config.getUInt64(prefix + "max_workers");
|
||||
|
||||
settings_common = Settings();
|
||||
if (config.has(prefix + "settings"))
|
||||
settings_common.loadSettingsFromConfig(prefix + "settings", config);
|
||||
|
||||
settings_common.prefer_localhost_replica = 0;
|
||||
|
||||
settings_pull = settings_common;
|
||||
if (config.has(prefix + "settings_pull"))
|
||||
settings_pull.loadSettingsFromConfig(prefix + "settings_pull", config);
|
||||
|
||||
settings_push = settings_common;
|
||||
if (config.has(prefix + "settings_push"))
|
||||
settings_push.loadSettingsFromConfig(prefix + "settings_push", config);
|
||||
|
||||
auto set_default_value = [] (auto && setting, auto && default_value)
|
||||
{
|
||||
setting = setting.changed ? setting.value : default_value;
|
||||
};
|
||||
|
||||
/// Override important settings
|
||||
settings_pull.readonly = 1;
|
||||
settings_pull.prefer_localhost_replica = false;
|
||||
settings_push.insert_distributed_sync = true;
|
||||
settings_push.prefer_localhost_replica = false;
|
||||
|
||||
set_default_value(settings_pull.load_balancing, LoadBalancing::NEAREST_HOSTNAME);
|
||||
set_default_value(settings_pull.max_threads, 1);
|
||||
set_default_value(settings_pull.max_block_size, 8192UL);
|
||||
set_default_value(settings_pull.preferred_block_size_bytes, 0);
|
||||
|
||||
set_default_value(settings_push.insert_distributed_timeout, 0);
|
||||
set_default_value(settings_push.replication_alter_partitions_sync, 2);
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
37
programs/copier/TaskShard.cpp
Normal file
37
programs/copier/TaskShard.cpp
Normal file
@ -0,0 +1,37 @@
|
||||
#include "TaskShard.h"
|
||||
|
||||
#include "TaskTable.h"
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
TaskShard::TaskShard(TaskTable & parent, const Cluster::ShardInfo & info_)
|
||||
: task_table(parent)
|
||||
, info(info_)
|
||||
{
|
||||
list_of_split_tables_on_shard.assign(task_table.number_of_splits, DatabaseAndTableName());
|
||||
}
|
||||
|
||||
UInt32 TaskShard::numberInCluster() const
|
||||
{
|
||||
return info.shard_num;
|
||||
}
|
||||
|
||||
UInt32 TaskShard::indexInCluster() const
|
||||
{
|
||||
return info.shard_num - 1;
|
||||
}
|
||||
|
||||
String DB::TaskShard::getDescription() const
|
||||
{
|
||||
return fmt::format("N{} (having a replica {}, pull table {} of cluster {}",
|
||||
numberInCluster(), getHostNameExample(), getQuotedTable(task_table.table_pull), task_table.cluster_pull_name);
|
||||
}
|
||||
|
||||
String DB::TaskShard::getHostNameExample() const
|
||||
{
|
||||
const auto & replicas = task_table.cluster_pull->getShardsAddresses().at(indexInCluster());
|
||||
return replicas.at(0).readableString();
|
||||
}
|
||||
|
||||
}
|
||||
56
programs/copier/TaskShard.h
Normal file
56
programs/copier/TaskShard.h
Normal file
@ -0,0 +1,56 @@
|
||||
#pragma once
|
||||
|
||||
#include "Aliases.h"
|
||||
#include "Internals.h"
|
||||
#include "ClusterPartition.h"
|
||||
#include "ShardPartition.h"
|
||||
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
struct TaskTable;
|
||||
|
||||
struct TaskShard
|
||||
{
|
||||
TaskShard(TaskTable & parent, const Cluster::ShardInfo & info_);
|
||||
|
||||
TaskTable & task_table;
|
||||
|
||||
Cluster::ShardInfo info;
|
||||
|
||||
UInt32 numberInCluster() const;
|
||||
|
||||
UInt32 indexInCluster() const;
|
||||
|
||||
String getDescription() const;
|
||||
|
||||
String getHostNameExample() const;
|
||||
|
||||
/// Used to sort clusters by their proximity
|
||||
ShardPriority priority;
|
||||
|
||||
/// Column with unique destination partitions (computed from engine_push_partition_key expr.) in the shard
|
||||
ColumnWithTypeAndName partition_key_column;
|
||||
|
||||
/// There is a task for each destination partition
|
||||
TasksPartition partition_tasks;
|
||||
|
||||
/// Which partitions have been checked for existence
|
||||
/// If some partition from this lists is exists, it is in partition_tasks
|
||||
std::set<String> checked_partitions;
|
||||
|
||||
/// Last CREATE TABLE query of the table of the shard
|
||||
ASTPtr current_pull_table_create_query;
|
||||
ASTPtr current_push_table_create_query;
|
||||
|
||||
/// Internal distributed tables
|
||||
DatabaseAndTableName table_read_shard;
|
||||
DatabaseAndTableName main_table_split_shard;
|
||||
ListOfDatabasesAndTableNames list_of_split_tables_on_shard;
|
||||
};
|
||||
|
||||
using TaskShardPtr = std::shared_ptr<TaskShard>;
|
||||
using TasksShard = std::vector<TaskShardPtr>;
|
||||
|
||||
}
|
||||
221
programs/copier/TaskTable.cpp
Normal file
221
programs/copier/TaskTable.cpp
Normal file
@ -0,0 +1,221 @@
|
||||
#include "TaskTable.h"
|
||||
|
||||
#include "ClusterPartition.h"
|
||||
#include "TaskCluster.h"
|
||||
|
||||
#include <Parsers/ASTFunction.h>
|
||||
|
||||
#include <boost/algorithm/string/join.hpp>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
namespace ErrorCodes
|
||||
{
|
||||
extern const int UNKNOWN_ELEMENT_IN_CONFIG;
|
||||
extern const int LOGICAL_ERROR;
|
||||
}
|
||||
|
||||
TaskTable::TaskTable(TaskCluster & parent, const Poco::Util::AbstractConfiguration & config,
|
||||
const String & prefix_, const String & table_key)
|
||||
: task_cluster(parent)
|
||||
{
|
||||
String table_prefix = prefix_ + "." + table_key + ".";
|
||||
|
||||
name_in_config = table_key;
|
||||
|
||||
number_of_splits = config.getUInt64(table_prefix + "number_of_splits", 3);
|
||||
|
||||
allow_to_copy_alias_and_materialized_columns = config.getBool(table_prefix + "allow_to_copy_alias_and_materialized_columns", false);
|
||||
allow_to_drop_target_partitions = config.getBool(table_prefix + "allow_to_drop_target_partitions", false);
|
||||
|
||||
cluster_pull_name = config.getString(table_prefix + "cluster_pull");
|
||||
cluster_push_name = config.getString(table_prefix + "cluster_push");
|
||||
|
||||
table_pull.first = config.getString(table_prefix + "database_pull");
|
||||
table_pull.second = config.getString(table_prefix + "table_pull");
|
||||
|
||||
table_push.first = config.getString(table_prefix + "database_push");
|
||||
table_push.second = config.getString(table_prefix + "table_push");
|
||||
|
||||
/// Used as node name in ZooKeeper
|
||||
table_id = escapeForFileName(cluster_push_name)
|
||||
+ "." + escapeForFileName(table_push.first)
|
||||
+ "." + escapeForFileName(table_push.second);
|
||||
|
||||
engine_push_str = config.getString(table_prefix + "engine", "rand()");
|
||||
|
||||
{
|
||||
ParserStorage parser_storage;
|
||||
engine_push_ast = parseQuery(parser_storage, engine_push_str, 0, DBMS_DEFAULT_MAX_PARSER_DEPTH);
|
||||
engine_push_partition_key_ast = extractPartitionKey(engine_push_ast);
|
||||
primary_key_comma_separated = boost::algorithm::join(extractPrimaryKeyColumnNames(engine_push_ast), ", ");
|
||||
is_replicated_table = isReplicatedTableEngine(engine_push_ast);
|
||||
}
|
||||
|
||||
sharding_key_str = config.getString(table_prefix + "sharding_key");
|
||||
|
||||
auxiliary_engine_split_asts.reserve(number_of_splits);
|
||||
{
|
||||
ParserExpressionWithOptionalAlias parser_expression(false);
|
||||
sharding_key_ast = parseQuery(parser_expression, sharding_key_str, 0, DBMS_DEFAULT_MAX_PARSER_DEPTH);
|
||||
main_engine_split_ast = createASTStorageDistributed(cluster_push_name, table_push.first, table_push.second,
|
||||
sharding_key_ast);
|
||||
|
||||
for (const auto piece_number : collections::range(0, number_of_splits))
|
||||
{
|
||||
auxiliary_engine_split_asts.emplace_back
|
||||
(
|
||||
createASTStorageDistributed(cluster_push_name, table_push.first,
|
||||
table_push.second + "_piece_" + toString(piece_number), sharding_key_ast)
|
||||
);
|
||||
}
|
||||
}
|
||||
|
||||
where_condition_str = config.getString(table_prefix + "where_condition", "");
|
||||
if (!where_condition_str.empty())
|
||||
{
|
||||
ParserExpressionWithOptionalAlias parser_expression(false);
|
||||
where_condition_ast = parseQuery(parser_expression, where_condition_str, 0, DBMS_DEFAULT_MAX_PARSER_DEPTH);
|
||||
|
||||
// Will use canonical expression form
|
||||
where_condition_str = queryToString(where_condition_ast);
|
||||
}
|
||||
|
||||
String enabled_partitions_prefix = table_prefix + "enabled_partitions";
|
||||
has_enabled_partitions = config.has(enabled_partitions_prefix);
|
||||
|
||||
if (has_enabled_partitions)
|
||||
{
|
||||
Strings keys;
|
||||
config.keys(enabled_partitions_prefix, keys);
|
||||
|
||||
if (keys.empty())
|
||||
{
|
||||
/// Parse list of partition from space-separated string
|
||||
String partitions_str = config.getString(table_prefix + "enabled_partitions");
|
||||
boost::trim_if(partitions_str, isWhitespaceASCII);
|
||||
boost::split(enabled_partitions, partitions_str, isWhitespaceASCII, boost::token_compress_on);
|
||||
}
|
||||
else
|
||||
{
|
||||
/// Parse sequence of <partition>...</partition>
|
||||
for (const String &key : keys)
|
||||
{
|
||||
if (!startsWith(key, "partition"))
|
||||
throw Exception("Unknown key " + key + " in " + enabled_partitions_prefix, ErrorCodes::UNKNOWN_ELEMENT_IN_CONFIG);
|
||||
|
||||
enabled_partitions.emplace_back(config.getString(enabled_partitions_prefix + "." + key));
|
||||
}
|
||||
}
|
||||
|
||||
std::copy(enabled_partitions.begin(), enabled_partitions.end(), std::inserter(enabled_partitions_set, enabled_partitions_set.begin()));
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
String TaskTable::getPartitionPath(const String & partition_name) const
|
||||
{
|
||||
return task_cluster.task_zookeeper_path // root
|
||||
+ "/tables/" + table_id // tables/dst_cluster.merge.hits
|
||||
+ "/" + escapeForFileName(partition_name); // 201701
|
||||
}
|
||||
|
||||
String TaskTable::getPartitionAttachIsActivePath(const String & partition_name) const
|
||||
{
|
||||
return getPartitionPath(partition_name) + "/attach_active";
|
||||
}
|
||||
|
||||
String TaskTable::getPartitionAttachIsDonePath(const String & partition_name) const
|
||||
{
|
||||
return getPartitionPath(partition_name) + "/attach_is_done";
|
||||
}
|
||||
|
||||
String TaskTable::getPartitionPiecePath(const String & partition_name, size_t piece_number) const
|
||||
{
|
||||
assert(piece_number < number_of_splits);
|
||||
return getPartitionPath(partition_name) + "/piece_" + toString(piece_number); // 1...number_of_splits
|
||||
}
|
||||
|
||||
String TaskTable::getCertainPartitionIsDirtyPath(const String &partition_name) const
|
||||
{
|
||||
return getPartitionPath(partition_name) + "/is_dirty";
|
||||
}
|
||||
|
||||
String TaskTable::getCertainPartitionPieceIsDirtyPath(const String & partition_name, const size_t piece_number) const
|
||||
{
|
||||
return getPartitionPiecePath(partition_name, piece_number) + "/is_dirty";
|
||||
}
|
||||
|
||||
String TaskTable::getCertainPartitionIsCleanedPath(const String & partition_name) const
|
||||
{
|
||||
return getCertainPartitionIsDirtyPath(partition_name) + "/cleaned";
|
||||
}
|
||||
|
||||
String TaskTable::getCertainPartitionPieceIsCleanedPath(const String & partition_name, const size_t piece_number) const
|
||||
{
|
||||
return getCertainPartitionPieceIsDirtyPath(partition_name, piece_number) + "/cleaned";
|
||||
}
|
||||
|
||||
String TaskTable::getCertainPartitionTaskStatusPath(const String & partition_name) const
|
||||
{
|
||||
return getPartitionPath(partition_name) + "/shards";
|
||||
}
|
||||
|
||||
String TaskTable::getCertainPartitionPieceTaskStatusPath(const String & partition_name, const size_t piece_number) const
|
||||
{
|
||||
return getPartitionPiecePath(partition_name, piece_number) + "/shards";
|
||||
}
|
||||
|
||||
bool TaskTable::isReplicatedTable() const
|
||||
{
|
||||
return is_replicated_table;
|
||||
}
|
||||
|
||||
String TaskTable::getStatusAllPartitionCount() const
|
||||
{
|
||||
return task_cluster.task_zookeeper_path + "/status/all_partitions_count";
|
||||
}
|
||||
|
||||
String TaskTable::getStatusProcessedPartitionsCount() const
|
||||
{
|
||||
return task_cluster.task_zookeeper_path + "/status/processed_partitions_count";
|
||||
}
|
||||
|
||||
ASTPtr TaskTable::rewriteReplicatedCreateQueryToPlain() const
|
||||
{
|
||||
ASTPtr prev_engine_push_ast = engine_push_ast->clone();
|
||||
|
||||
auto & new_storage_ast = prev_engine_push_ast->as<ASTStorage &>();
|
||||
auto & new_engine_ast = new_storage_ast.engine->as<ASTFunction &>();
|
||||
|
||||
/// Remove "Replicated" from name
|
||||
new_engine_ast.name = new_engine_ast.name.substr(10);
|
||||
|
||||
if (new_engine_ast.arguments)
|
||||
{
|
||||
auto & replicated_table_arguments = new_engine_ast.arguments->children;
|
||||
|
||||
|
||||
/// In some cases of Atomic database engine usage ReplicatedMergeTree tables
|
||||
/// could be created without arguments.
|
||||
if (!replicated_table_arguments.empty())
|
||||
{
|
||||
/// Delete first two arguments of Replicated...MergeTree() table.
|
||||
replicated_table_arguments.erase(replicated_table_arguments.begin());
|
||||
replicated_table_arguments.erase(replicated_table_arguments.begin());
|
||||
}
|
||||
}
|
||||
|
||||
return new_storage_ast.clone();
|
||||
}
|
||||
|
||||
ClusterPartition & TaskTable::getClusterPartition(const String & partition_name)
|
||||
{
|
||||
auto it = cluster_partitions.find(partition_name);
|
||||
if (it == cluster_partitions.end())
|
||||
throw Exception("There are no cluster partition " + partition_name + " in " + table_id,
|
||||
ErrorCodes::LOGICAL_ERROR);
|
||||
return it->second;
|
||||
}
|
||||
|
||||
}
|
||||
173
programs/copier/TaskTable.h
Normal file
173
programs/copier/TaskTable.h
Normal file
@ -0,0 +1,173 @@
|
||||
#pragma once
|
||||
|
||||
#include "Aliases.h"
|
||||
#include "TaskShard.h"
|
||||
|
||||
|
||||
namespace DB
|
||||
{
|
||||
|
||||
struct ClusterPartition;
|
||||
struct TaskCluster;
|
||||
|
||||
struct TaskTable
|
||||
{
|
||||
TaskTable(TaskCluster & parent, const Poco::Util::AbstractConfiguration & config, const String & prefix, const String & table_key);
|
||||
|
||||
TaskCluster & task_cluster;
|
||||
|
||||
/// These functions used in checkPartitionIsDone() or checkPartitionPieceIsDone()
|
||||
/// They are implemented here not to call task_table.tasks_shard[partition_name].second.pieces[current_piece_number] etc.
|
||||
|
||||
String getPartitionPath(const String & partition_name) const;
|
||||
|
||||
String getPartitionAttachIsActivePath(const String & partition_name) const;
|
||||
|
||||
String getPartitionAttachIsDonePath(const String & partition_name) const;
|
||||
|
||||
String getPartitionPiecePath(const String & partition_name, size_t piece_number) const;
|
||||
|
||||
String getCertainPartitionIsDirtyPath(const String & partition_name) const;
|
||||
|
||||
String getCertainPartitionPieceIsDirtyPath(const String & partition_name, size_t piece_number) const;
|
||||
|
||||
String getCertainPartitionIsCleanedPath(const String & partition_name) const;
|
||||
|
||||
String getCertainPartitionPieceIsCleanedPath(const String & partition_name, size_t piece_number) const;
|
||||
|
||||
String getCertainPartitionTaskStatusPath(const String & partition_name) const;
|
||||
|
||||
String getCertainPartitionPieceTaskStatusPath(const String & partition_name, size_t piece_number) const;
|
||||
|
||||
bool isReplicatedTable() const;
|
||||
|
||||
/// These nodes are used for check-status option
|
||||
String getStatusAllPartitionCount() const;
|
||||
String getStatusProcessedPartitionsCount() const;
|
||||
|
||||
/// Partitions will be split into number-of-splits pieces.
|
||||
/// Each piece will be copied independently. (10 by default)
|
||||
size_t number_of_splits;
|
||||
|
||||
bool allow_to_copy_alias_and_materialized_columns{false};
|
||||
bool allow_to_drop_target_partitions{false};
|
||||
|
||||
String name_in_config;
|
||||
|
||||
/// Used as task ID
|
||||
String table_id;
|
||||
|
||||
/// Column names in primary key
|
||||
String primary_key_comma_separated;
|
||||
|
||||
/// Source cluster and table
|
||||
String cluster_pull_name;
|
||||
DatabaseAndTableName table_pull;
|
||||
|
||||
/// Destination cluster and table
|
||||
String cluster_push_name;
|
||||
DatabaseAndTableName table_push;
|
||||
|
||||
/// Storage of destination table
|
||||
/// (tables that are stored on each shard of target cluster)
|
||||
String engine_push_str;
|
||||
ASTPtr engine_push_ast;
|
||||
ASTPtr engine_push_partition_key_ast;
|
||||
|
||||
/// First argument of Replicated...MergeTree()
|
||||
String engine_push_zk_path;
|
||||
bool is_replicated_table;
|
||||
|
||||
ASTPtr rewriteReplicatedCreateQueryToPlain() const;
|
||||
|
||||
/*
|
||||
* A Distributed table definition used to split data
|
||||
* Distributed table will be created on each shard of default
|
||||
* cluster to perform data copying and resharding
|
||||
* */
|
||||
String sharding_key_str;
|
||||
ASTPtr sharding_key_ast;
|
||||
ASTPtr main_engine_split_ast;
|
||||
|
||||
/*
|
||||
* To copy partition piece form one cluster to another we have to use Distributed table.
|
||||
* In case of usage separate table (engine_push) for each partition piece,
|
||||
* we have to use many Distributed tables.
|
||||
* */
|
||||
ASTs auxiliary_engine_split_asts;
|
||||
|
||||
/// Additional WHERE expression to filter input data
|
||||
String where_condition_str;
|
||||
ASTPtr where_condition_ast;
|
||||
|
||||
/// Resolved clusters
|
||||
ClusterPtr cluster_pull;
|
||||
ClusterPtr cluster_push;
|
||||
|
||||
/// Filter partitions that should be copied
|
||||
bool has_enabled_partitions = false;
|
||||
Strings enabled_partitions;
|
||||
NameSet enabled_partitions_set;
|
||||
|
||||
/**
|
||||
* Prioritized list of shards
|
||||
* all_shards contains information about all shards in the table.
|
||||
* So we have to check whether particular shard have current partition or not while processing.
|
||||
*/
|
||||
TasksShard all_shards;
|
||||
TasksShard local_shards;
|
||||
|
||||
/// All partitions of the current table.
|
||||
ClusterPartitions cluster_partitions;
|
||||
NameSet finished_cluster_partitions;
|
||||
|
||||
/// Partition names to process in user-specified order
|
||||
Strings ordered_partition_names;
|
||||
|
||||
ClusterPartition & getClusterPartition(const String & partition_name);
|
||||
|
||||
Stopwatch watch;
|
||||
UInt64 bytes_copied = 0;
|
||||
UInt64 rows_copied = 0;
|
||||
|
||||
template <typename RandomEngine>
|
||||
void initShards(RandomEngine &&random_engine);
|
||||
};
|
||||
|
||||
using TasksTable = std::list<TaskTable>;
|
||||
|
||||
|
||||
template<typename RandomEngine>
|
||||
inline void TaskTable::initShards(RandomEngine && random_engine)
|
||||
{
|
||||
const String & fqdn_name = getFQDNOrHostName();
|
||||
std::uniform_int_distribution<uint8_t> get_urand(0, std::numeric_limits<UInt8>::max());
|
||||
|
||||
// Compute the priority
|
||||
for (const auto & shard_info : cluster_pull->getShardsInfo())
|
||||
{
|
||||
TaskShardPtr task_shard = std::make_shared<TaskShard>(*this, shard_info);
|
||||
const auto & replicas = cluster_pull->getShardsAddresses().at(task_shard->indexInCluster());
|
||||
task_shard->priority = getReplicasPriority(replicas, fqdn_name, get_urand(random_engine));
|
||||
|
||||
all_shards.emplace_back(task_shard);
|
||||
}
|
||||
|
||||
// Sort by priority
|
||||
std::sort(all_shards.begin(), all_shards.end(),
|
||||
[](const TaskShardPtr & lhs, const TaskShardPtr & rhs)
|
||||
{
|
||||
return ShardPriority::greaterPriority(lhs->priority, rhs->priority);
|
||||
});
|
||||
|
||||
// Cut local shards
|
||||
auto it_first_remote = std::lower_bound(all_shards.begin(), all_shards.end(), 1,
|
||||
[](const TaskShardPtr & lhs, UInt8 is_remote)
|
||||
{
|
||||
return lhs->priority.is_remote < is_remote;
|
||||
});
|
||||
|
||||
local_shards.assign(all_shards.begin(), it_first_remote);
|
||||
}
|
||||
|
||||
}
|
||||
@ -1,434 +0,0 @@
|
||||
#pragma once
|
||||
|
||||
#include "Aliases.h"
|
||||
#include "Internals.h"
|
||||
#include "ClusterPartition.h"
|
||||
|
||||
#include <Core/Defines.h>
|
||||
#include <Parsers/ASTFunction.h>
|
||||
|
||||
#include <base/map.h>
|
||||
#include <boost/algorithm/string/join.hpp>
|
||||
|
||||
|
||||
namespace DB
|
||||
{
|
||||
namespace ErrorCodes
|
||||
{
|
||||
extern const int UNKNOWN_ELEMENT_IN_CONFIG;
|
||||
extern const int LOGICAL_ERROR;
|
||||
}
|
||||
|
||||
struct TaskShard;
|
||||
|
||||
struct TaskTable
|
||||
{
|
||||
TaskTable(TaskCluster & parent, const Poco::Util::AbstractConfiguration & config, const String & prefix,
|
||||
const String & table_key);
|
||||
|
||||
TaskCluster & task_cluster;
|
||||
|
||||
/// These functions used in checkPartitionIsDone() or checkPartitionPieceIsDone()
|
||||
/// They are implemented here not to call task_table.tasks_shard[partition_name].second.pieces[current_piece_number] etc.
|
||||
|
||||
String getPartitionPath(const String & partition_name) const;
|
||||
|
||||
String getPartitionAttachIsActivePath(const String & partition_name) const;
|
||||
|
||||
String getPartitionAttachIsDonePath(const String & partition_name) const;
|
||||
|
||||
String getPartitionPiecePath(const String & partition_name, size_t piece_number) const;
|
||||
|
||||
String getCertainPartitionIsDirtyPath(const String & partition_name) const;
|
||||
|
||||
String getCertainPartitionPieceIsDirtyPath(const String & partition_name, size_t piece_number) const;
|
||||
|
||||
String getCertainPartitionIsCleanedPath(const String & partition_name) const;
|
||||
|
||||
String getCertainPartitionPieceIsCleanedPath(const String & partition_name, size_t piece_number) const;
|
||||
|
||||
String getCertainPartitionTaskStatusPath(const String & partition_name) const;
|
||||
|
||||
String getCertainPartitionPieceTaskStatusPath(const String & partition_name, size_t piece_number) const;
|
||||
|
||||
bool isReplicatedTable() const { return is_replicated_table; }
|
||||
|
||||
/// These nodes are used for check-status option
|
||||
String getStatusAllPartitionCount() const;
|
||||
String getStatusProcessedPartitionsCount() const;
|
||||
|
||||
/// Partitions will be split into number-of-splits pieces.
|
||||
/// Each piece will be copied independently. (10 by default)
|
||||
size_t number_of_splits;
|
||||
|
||||
bool allow_to_copy_alias_and_materialized_columns{false};
|
||||
bool allow_to_drop_target_partitions{false};
|
||||
|
||||
String name_in_config;
|
||||
|
||||
/// Used as task ID
|
||||
String table_id;
|
||||
|
||||
/// Column names in primary key
|
||||
String primary_key_comma_separated;
|
||||
|
||||
/// Source cluster and table
|
||||
String cluster_pull_name;
|
||||
DatabaseAndTableName table_pull;
|
||||
|
||||
/// Destination cluster and table
|
||||
String cluster_push_name;
|
||||
DatabaseAndTableName table_push;
|
||||
|
||||
/// Storage of destination table
|
||||
/// (tables that are stored on each shard of target cluster)
|
||||
String engine_push_str;
|
||||
ASTPtr engine_push_ast;
|
||||
ASTPtr engine_push_partition_key_ast;
|
||||
|
||||
/// First argument of Replicated...MergeTree()
|
||||
String engine_push_zk_path;
|
||||
bool is_replicated_table;
|
||||
|
||||
ASTPtr rewriteReplicatedCreateQueryToPlain() const;
|
||||
|
||||
/*
|
||||
* A Distributed table definition used to split data
|
||||
* Distributed table will be created on each shard of default
|
||||
* cluster to perform data copying and resharding
|
||||
* */
|
||||
String sharding_key_str;
|
||||
ASTPtr sharding_key_ast;
|
||||
ASTPtr main_engine_split_ast;
|
||||
|
||||
/*
|
||||
* To copy partition piece form one cluster to another we have to use Distributed table.
|
||||
* In case of usage separate table (engine_push) for each partition piece,
|
||||
* we have to use many Distributed tables.
|
||||
* */
|
||||
ASTs auxiliary_engine_split_asts;
|
||||
|
||||
/// Additional WHERE expression to filter input data
|
||||
String where_condition_str;
|
||||
ASTPtr where_condition_ast;
|
||||
|
||||
/// Resolved clusters
|
||||
ClusterPtr cluster_pull;
|
||||
ClusterPtr cluster_push;
|
||||
|
||||
/// Filter partitions that should be copied
|
||||
bool has_enabled_partitions = false;
|
||||
Strings enabled_partitions;
|
||||
NameSet enabled_partitions_set;
|
||||
|
||||
/**
|
||||
* Prioritized list of shards
|
||||
* all_shards contains information about all shards in the table.
|
||||
* So we have to check whether particular shard have current partition or not while processing.
|
||||
*/
|
||||
TasksShard all_shards;
|
||||
TasksShard local_shards;
|
||||
|
||||
/// All partitions of the current table.
|
||||
ClusterPartitions cluster_partitions;
|
||||
NameSet finished_cluster_partitions;
|
||||
|
||||
/// Partition names to process in user-specified order
|
||||
Strings ordered_partition_names;
|
||||
|
||||
ClusterPartition & getClusterPartition(const String & partition_name)
|
||||
{
|
||||
auto it = cluster_partitions.find(partition_name);
|
||||
if (it == cluster_partitions.end())
|
||||
throw Exception("There are no cluster partition " + partition_name + " in " + table_id,
|
||||
ErrorCodes::LOGICAL_ERROR);
|
||||
return it->second;
|
||||
}
|
||||
|
||||
Stopwatch watch;
|
||||
UInt64 bytes_copied = 0;
|
||||
UInt64 rows_copied = 0;
|
||||
|
||||
template <typename RandomEngine>
|
||||
void initShards(RandomEngine &&random_engine);
|
||||
};
|
||||
|
||||
|
||||
struct TaskShard
|
||||
{
|
||||
TaskShard(TaskTable & parent, const ShardInfo & info_) : task_table(parent), info(info_)
|
||||
{
|
||||
list_of_split_tables_on_shard.assign(task_table.number_of_splits, DatabaseAndTableName());
|
||||
}
|
||||
|
||||
TaskTable & task_table;
|
||||
|
||||
ShardInfo info;
|
||||
|
||||
UInt32 numberInCluster() const { return info.shard_num; }
|
||||
|
||||
UInt32 indexInCluster() const { return info.shard_num - 1; }
|
||||
|
||||
String getDescription() const;
|
||||
|
||||
String getHostNameExample() const;
|
||||
|
||||
/// Used to sort clusters by their proximity
|
||||
ShardPriority priority;
|
||||
|
||||
/// Column with unique destination partitions (computed from engine_push_partition_key expr.) in the shard
|
||||
ColumnWithTypeAndName partition_key_column;
|
||||
|
||||
/// There is a task for each destination partition
|
||||
TasksPartition partition_tasks;
|
||||
|
||||
/// Which partitions have been checked for existence
|
||||
/// If some partition from this lists is exists, it is in partition_tasks
|
||||
std::set<String> checked_partitions;
|
||||
|
||||
/// Last CREATE TABLE query of the table of the shard
|
||||
ASTPtr current_pull_table_create_query;
|
||||
ASTPtr current_push_table_create_query;
|
||||
|
||||
/// Internal distributed tables
|
||||
DatabaseAndTableName table_read_shard;
|
||||
DatabaseAndTableName main_table_split_shard;
|
||||
ListOfDatabasesAndTableNames list_of_split_tables_on_shard;
|
||||
};
|
||||
|
||||
|
||||
inline String TaskTable::getPartitionPath(const String & partition_name) const
|
||||
{
|
||||
return task_cluster.task_zookeeper_path // root
|
||||
+ "/tables/" + table_id // tables/dst_cluster.merge.hits
|
||||
+ "/" + escapeForFileName(partition_name); // 201701
|
||||
}
|
||||
|
||||
inline String TaskTable::getPartitionAttachIsActivePath(const String & partition_name) const
|
||||
{
|
||||
return getPartitionPath(partition_name) + "/attach_active";
|
||||
}
|
||||
|
||||
inline String TaskTable::getPartitionAttachIsDonePath(const String & partition_name) const
|
||||
{
|
||||
return getPartitionPath(partition_name) + "/attach_is_done";
|
||||
}
|
||||
|
||||
inline String TaskTable::getPartitionPiecePath(const String & partition_name, size_t piece_number) const
|
||||
{
|
||||
assert(piece_number < number_of_splits);
|
||||
return getPartitionPath(partition_name) + "/piece_" + toString(piece_number); // 1...number_of_splits
|
||||
}
|
||||
|
||||
inline String TaskTable::getCertainPartitionIsDirtyPath(const String &partition_name) const
|
||||
{
|
||||
return getPartitionPath(partition_name) + "/is_dirty";
|
||||
}
|
||||
|
||||
inline String TaskTable::getCertainPartitionPieceIsDirtyPath(const String & partition_name, const size_t piece_number) const
|
||||
{
|
||||
return getPartitionPiecePath(partition_name, piece_number) + "/is_dirty";
|
||||
}
|
||||
|
||||
inline String TaskTable::getCertainPartitionIsCleanedPath(const String & partition_name) const
|
||||
{
|
||||
return getCertainPartitionIsDirtyPath(partition_name) + "/cleaned";
|
||||
}
|
||||
|
||||
inline String TaskTable::getCertainPartitionPieceIsCleanedPath(const String & partition_name, const size_t piece_number) const
|
||||
{
|
||||
return getCertainPartitionPieceIsDirtyPath(partition_name, piece_number) + "/cleaned";
|
||||
}
|
||||
|
||||
inline String TaskTable::getCertainPartitionTaskStatusPath(const String & partition_name) const
|
||||
{
|
||||
return getPartitionPath(partition_name) + "/shards";
|
||||
}
|
||||
|
||||
inline String TaskTable::getCertainPartitionPieceTaskStatusPath(const String & partition_name, const size_t piece_number) const
|
||||
{
|
||||
return getPartitionPiecePath(partition_name, piece_number) + "/shards";
|
||||
}
|
||||
|
||||
inline String TaskTable::getStatusAllPartitionCount() const
|
||||
{
|
||||
return task_cluster.task_zookeeper_path + "/status/all_partitions_count";
|
||||
}
|
||||
|
||||
inline String TaskTable::getStatusProcessedPartitionsCount() const
|
||||
{
|
||||
return task_cluster.task_zookeeper_path + "/status/processed_partitions_count";
|
||||
}
|
||||
|
||||
inline TaskTable::TaskTable(TaskCluster & parent, const Poco::Util::AbstractConfiguration & config,
|
||||
const String & prefix_, const String & table_key)
|
||||
: task_cluster(parent)
|
||||
{
|
||||
String table_prefix = prefix_ + "." + table_key + ".";
|
||||
|
||||
name_in_config = table_key;
|
||||
|
||||
number_of_splits = config.getUInt64(table_prefix + "number_of_splits", 3);
|
||||
|
||||
allow_to_copy_alias_and_materialized_columns = config.getBool(table_prefix + "allow_to_copy_alias_and_materialized_columns", false);
|
||||
allow_to_drop_target_partitions = config.getBool(table_prefix + "allow_to_drop_target_partitions", false);
|
||||
|
||||
cluster_pull_name = config.getString(table_prefix + "cluster_pull");
|
||||
cluster_push_name = config.getString(table_prefix + "cluster_push");
|
||||
|
||||
table_pull.first = config.getString(table_prefix + "database_pull");
|
||||
table_pull.second = config.getString(table_prefix + "table_pull");
|
||||
|
||||
table_push.first = config.getString(table_prefix + "database_push");
|
||||
table_push.second = config.getString(table_prefix + "table_push");
|
||||
|
||||
/// Used as node name in ZooKeeper
|
||||
table_id = escapeForFileName(cluster_push_name)
|
||||
+ "." + escapeForFileName(table_push.first)
|
||||
+ "." + escapeForFileName(table_push.second);
|
||||
|
||||
engine_push_str = config.getString(table_prefix + "engine", "rand()");
|
||||
|
||||
{
|
||||
ParserStorage parser_storage;
|
||||
engine_push_ast = parseQuery(parser_storage, engine_push_str, 0, DBMS_DEFAULT_MAX_PARSER_DEPTH);
|
||||
engine_push_partition_key_ast = extractPartitionKey(engine_push_ast);
|
||||
primary_key_comma_separated = boost::algorithm::join(extractPrimaryKeyColumnNames(engine_push_ast), ", ");
|
||||
is_replicated_table = isReplicatedTableEngine(engine_push_ast);
|
||||
}
|
||||
|
||||
sharding_key_str = config.getString(table_prefix + "sharding_key");
|
||||
|
||||
auxiliary_engine_split_asts.reserve(number_of_splits);
|
||||
{
|
||||
ParserExpressionWithOptionalAlias parser_expression(false);
|
||||
sharding_key_ast = parseQuery(parser_expression, sharding_key_str, 0, DBMS_DEFAULT_MAX_PARSER_DEPTH);
|
||||
main_engine_split_ast = createASTStorageDistributed(cluster_push_name, table_push.first, table_push.second,
|
||||
sharding_key_ast);
|
||||
|
||||
for (const auto piece_number : collections::range(0, number_of_splits))
|
||||
{
|
||||
auxiliary_engine_split_asts.emplace_back
|
||||
(
|
||||
createASTStorageDistributed(cluster_push_name, table_push.first,
|
||||
table_push.second + "_piece_" + toString(piece_number), sharding_key_ast)
|
||||
);
|
||||
}
|
||||
}
|
||||
|
||||
where_condition_str = config.getString(table_prefix + "where_condition", "");
|
||||
if (!where_condition_str.empty())
|
||||
{
|
||||
ParserExpressionWithOptionalAlias parser_expression(false);
|
||||
where_condition_ast = parseQuery(parser_expression, where_condition_str, 0, DBMS_DEFAULT_MAX_PARSER_DEPTH);
|
||||
|
||||
// Will use canonical expression form
|
||||
where_condition_str = queryToString(where_condition_ast);
|
||||
}
|
||||
|
||||
String enabled_partitions_prefix = table_prefix + "enabled_partitions";
|
||||
has_enabled_partitions = config.has(enabled_partitions_prefix);
|
||||
|
||||
if (has_enabled_partitions)
|
||||
{
|
||||
Strings keys;
|
||||
config.keys(enabled_partitions_prefix, keys);
|
||||
|
||||
if (keys.empty())
|
||||
{
|
||||
/// Parse list of partition from space-separated string
|
||||
String partitions_str = config.getString(table_prefix + "enabled_partitions");
|
||||
boost::trim_if(partitions_str, isWhitespaceASCII);
|
||||
boost::split(enabled_partitions, partitions_str, isWhitespaceASCII, boost::token_compress_on);
|
||||
}
|
||||
else
|
||||
{
|
||||
/// Parse sequence of <partition>...</partition>
|
||||
for (const String &key : keys)
|
||||
{
|
||||
if (!startsWith(key, "partition"))
|
||||
throw Exception("Unknown key " + key + " in " + enabled_partitions_prefix, ErrorCodes::UNKNOWN_ELEMENT_IN_CONFIG);
|
||||
|
||||
enabled_partitions.emplace_back(config.getString(enabled_partitions_prefix + "." + key));
|
||||
}
|
||||
}
|
||||
|
||||
std::copy(enabled_partitions.begin(), enabled_partitions.end(), std::inserter(enabled_partitions_set, enabled_partitions_set.begin()));
|
||||
}
|
||||
}
|
||||
|
||||
template<typename RandomEngine>
|
||||
inline void TaskTable::initShards(RandomEngine && random_engine)
|
||||
{
|
||||
const String & fqdn_name = getFQDNOrHostName();
|
||||
std::uniform_int_distribution<UInt8> get_urand(0, std::numeric_limits<UInt8>::max());
|
||||
|
||||
// Compute the priority
|
||||
for (const auto & shard_info : cluster_pull->getShardsInfo())
|
||||
{
|
||||
TaskShardPtr task_shard = std::make_shared<TaskShard>(*this, shard_info);
|
||||
const auto & replicas = cluster_pull->getShardsAddresses().at(task_shard->indexInCluster());
|
||||
task_shard->priority = getReplicasPriority(replicas, fqdn_name, get_urand(random_engine));
|
||||
|
||||
all_shards.emplace_back(task_shard);
|
||||
}
|
||||
|
||||
// Sort by priority
|
||||
std::sort(all_shards.begin(), all_shards.end(),
|
||||
[](const TaskShardPtr & lhs, const TaskShardPtr & rhs)
|
||||
{
|
||||
return ShardPriority::greaterPriority(lhs->priority, rhs->priority);
|
||||
});
|
||||
|
||||
// Cut local shards
|
||||
auto it_first_remote = std::lower_bound(all_shards.begin(), all_shards.end(), 1,
|
||||
[](const TaskShardPtr & lhs, UInt8 is_remote)
|
||||
{
|
||||
return lhs->priority.is_remote < is_remote;
|
||||
});
|
||||
|
||||
local_shards.assign(all_shards.begin(), it_first_remote);
|
||||
}
|
||||
|
||||
inline ASTPtr TaskTable::rewriteReplicatedCreateQueryToPlain() const
|
||||
{
|
||||
ASTPtr prev_engine_push_ast = engine_push_ast->clone();
|
||||
|
||||
auto & new_storage_ast = prev_engine_push_ast->as<ASTStorage &>();
|
||||
auto & new_engine_ast = new_storage_ast.engine->as<ASTFunction &>();
|
||||
|
||||
/// Remove "Replicated" from name
|
||||
new_engine_ast.name = new_engine_ast.name.substr(10);
|
||||
|
||||
if (new_engine_ast.arguments)
|
||||
{
|
||||
auto & replicated_table_arguments = new_engine_ast.arguments->children;
|
||||
|
||||
|
||||
/// In some cases of Atomic database engine usage ReplicatedMergeTree tables
|
||||
/// could be created without arguments.
|
||||
if (!replicated_table_arguments.empty())
|
||||
{
|
||||
/// Delete first two arguments of Replicated...MergeTree() table.
|
||||
replicated_table_arguments.erase(replicated_table_arguments.begin());
|
||||
replicated_table_arguments.erase(replicated_table_arguments.begin());
|
||||
}
|
||||
}
|
||||
|
||||
return new_storage_ast.clone();
|
||||
}
|
||||
|
||||
|
||||
inline String DB::TaskShard::getDescription() const
|
||||
{
|
||||
return fmt::format("N{} (having a replica {}, pull table {} of cluster {}",
|
||||
numberInCluster(), getHostNameExample(), getQuotedTable(task_table.table_pull), task_table.cluster_pull_name);
|
||||
}
|
||||
|
||||
inline String DB::TaskShard::getHostNameExample() const
|
||||
{
|
||||
const auto & replicas = task_table.cluster_pull->getShardsAddresses().at(indexInCluster());
|
||||
return replicas.at(0).readableString();
|
||||
}
|
||||
|
||||
}
|
||||
@ -10,23 +10,34 @@ namespace DB
|
||||
namespace
|
||||
{
|
||||
|
||||
struct QueryTreeNodeHash
|
||||
struct QueryTreeNodeWithHash
|
||||
{
|
||||
size_t operator()(const IQueryTreeNode * node) const
|
||||
explicit QueryTreeNodeWithHash(const IQueryTreeNode * node_)
|
||||
: node(node_)
|
||||
, hash(node->getTreeHash().first)
|
||||
{}
|
||||
|
||||
const IQueryTreeNode * node = nullptr;
|
||||
size_t hash = 0;
|
||||
};
|
||||
|
||||
struct QueryTreeNodeWithHashHash
|
||||
{
|
||||
size_t operator()(const QueryTreeNodeWithHash & node_with_hash) const
|
||||
{
|
||||
return node->getTreeHash().first;
|
||||
return node_with_hash.hash;
|
||||
}
|
||||
};
|
||||
|
||||
struct QueryTreeNodeEqualTo
|
||||
struct QueryTreeNodeWithHashEqualTo
|
||||
{
|
||||
size_t operator()(const IQueryTreeNode * lhs_node, const IQueryTreeNode * rhs_node) const
|
||||
bool operator()(const QueryTreeNodeWithHash & lhs_node, const QueryTreeNodeWithHash & rhs_node) const
|
||||
{
|
||||
return lhs_node->isEqual(*rhs_node);
|
||||
return lhs_node.hash == rhs_node.hash && lhs_node.node->isEqual(*rhs_node.node);
|
||||
}
|
||||
};
|
||||
|
||||
using QueryTreeNodeSet = std::unordered_set<const IQueryTreeNode *, QueryTreeNodeHash, QueryTreeNodeEqualTo>;
|
||||
using QueryTreeNodeWithHashSet = std::unordered_set<QueryTreeNodeWithHash, QueryTreeNodeWithHashHash, QueryTreeNodeWithHashEqualTo>;
|
||||
|
||||
class OrderByLimitByDuplicateEliminationVisitor : public InDepthQueryTreeVisitor<OrderByLimitByDuplicateEliminationVisitor>
|
||||
{
|
||||
@ -82,7 +93,7 @@ public:
|
||||
}
|
||||
|
||||
private:
|
||||
QueryTreeNodeSet unique_expressions_nodes_set;
|
||||
QueryTreeNodeWithHashSet unique_expressions_nodes_set;
|
||||
};
|
||||
|
||||
}
|
||||
|
||||
@ -2,6 +2,7 @@
|
||||
#include <arpa/inet.h>
|
||||
#include <sys/select.h>
|
||||
#include <Common/Exception.h>
|
||||
#include <Common/logger_useful.h>
|
||||
#include "ares.h"
|
||||
#include "netdb.h"
|
||||
|
||||
@ -40,6 +41,8 @@ namespace DB
|
||||
}
|
||||
}
|
||||
|
||||
std::mutex CaresPTRResolver::mutex;
|
||||
|
||||
CaresPTRResolver::CaresPTRResolver(CaresPTRResolver::provider_token) : channel(nullptr)
|
||||
{
|
||||
/*
|
||||
@ -73,6 +76,8 @@ namespace DB
|
||||
|
||||
std::unordered_set<std::string> CaresPTRResolver::resolve(const std::string & ip)
|
||||
{
|
||||
std::lock_guard guard(mutex);
|
||||
|
||||
std::unordered_set<std::string> ptr_records;
|
||||
|
||||
resolve(ip, ptr_records);
|
||||
@ -83,6 +88,8 @@ namespace DB
|
||||
|
||||
std::unordered_set<std::string> CaresPTRResolver::resolve_v6(const std::string & ip)
|
||||
{
|
||||
std::lock_guard guard(mutex);
|
||||
|
||||
std::unordered_set<std::string> ptr_records;
|
||||
|
||||
resolve_v6(ip, ptr_records);
|
||||
@ -110,23 +117,83 @@ namespace DB
|
||||
|
||||
void CaresPTRResolver::wait()
|
||||
{
|
||||
timeval * tvp, tv;
|
||||
fd_set read_fds;
|
||||
fd_set write_fds;
|
||||
int nfds;
|
||||
int sockets[ARES_GETSOCK_MAXNUM];
|
||||
pollfd pollfd[ARES_GETSOCK_MAXNUM];
|
||||
|
||||
for (;;)
|
||||
while (true)
|
||||
{
|
||||
FD_ZERO(&read_fds);
|
||||
FD_ZERO(&write_fds);
|
||||
nfds = ares_fds(channel, &read_fds,&write_fds);
|
||||
if (nfds == 0)
|
||||
auto readable_sockets = get_readable_sockets(sockets, pollfd);
|
||||
auto timeout = calculate_timeout();
|
||||
|
||||
int number_of_fds_ready = 0;
|
||||
if (!readable_sockets.empty())
|
||||
{
|
||||
number_of_fds_ready = poll(readable_sockets.data(), readable_sockets.size(), static_cast<int>(timeout));
|
||||
}
|
||||
|
||||
if (number_of_fds_ready > 0)
|
||||
{
|
||||
process_readable_sockets(readable_sockets);
|
||||
}
|
||||
else
|
||||
{
|
||||
process_possible_timeout();
|
||||
break;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
std::span<pollfd> CaresPTRResolver::get_readable_sockets(int * sockets, pollfd * pollfd)
|
||||
{
|
||||
int sockets_bitmask = ares_getsock(channel, sockets, ARES_GETSOCK_MAXNUM);
|
||||
|
||||
int number_of_sockets_to_poll = 0;
|
||||
|
||||
for (int i = 0; i < ARES_GETSOCK_MAXNUM; i++, number_of_sockets_to_poll++)
|
||||
{
|
||||
pollfd[i].events = 0;
|
||||
pollfd[i].revents = 0;
|
||||
|
||||
if (ARES_GETSOCK_READABLE(sockets_bitmask, i))
|
||||
{
|
||||
pollfd[i].fd = sockets[i];
|
||||
pollfd[i].events = POLLIN;
|
||||
}
|
||||
else
|
||||
{
|
||||
break;
|
||||
}
|
||||
tvp = ares_timeout(channel, nullptr, &tv);
|
||||
select(nfds, &read_fds, &write_fds, nullptr, tvp);
|
||||
ares_process(channel, &read_fds, &write_fds);
|
||||
}
|
||||
|
||||
return std::span<struct pollfd>(pollfd, number_of_sockets_to_poll);
|
||||
}
|
||||
|
||||
int64_t CaresPTRResolver::calculate_timeout()
|
||||
{
|
||||
timeval tv;
|
||||
if (auto * tvp = ares_timeout(channel, nullptr, &tv))
|
||||
{
|
||||
auto timeout = tvp->tv_sec * 1000 + tvp->tv_usec / 1000;
|
||||
|
||||
return timeout;
|
||||
}
|
||||
|
||||
return 0;
|
||||
}
|
||||
|
||||
void CaresPTRResolver::process_possible_timeout()
|
||||
{
|
||||
/* Call ares_process() unconditonally here, even if we simply timed out
|
||||
above, as otherwise the ares name resolve won't timeout! */
|
||||
ares_process_fd(channel, ARES_SOCKET_BAD, ARES_SOCKET_BAD);
|
||||
}
|
||||
|
||||
void CaresPTRResolver::process_readable_sockets(std::span<pollfd> readable_sockets)
|
||||
{
|
||||
for (auto readable_socket : readable_sockets)
|
||||
{
|
||||
auto fd = readable_socket.revents & POLLIN ? readable_socket.fd : ARES_SOCKET_BAD;
|
||||
ares_process_fd(channel, fd, ARES_SOCKET_BAD);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
@ -1,5 +1,8 @@
|
||||
#pragma once
|
||||
|
||||
#include <span>
|
||||
#include <poll.h>
|
||||
#include <mutex>
|
||||
#include "DNSPTRResolver.h"
|
||||
|
||||
using ares_channel = struct ares_channeldata *;
|
||||
@ -20,7 +23,6 @@ namespace DB
|
||||
* Allow only DNSPTRProvider to instantiate this class
|
||||
* */
|
||||
struct provider_token {};
|
||||
|
||||
public:
|
||||
explicit CaresPTRResolver(provider_token);
|
||||
~CaresPTRResolver() override;
|
||||
@ -36,7 +38,17 @@ namespace DB
|
||||
|
||||
void resolve_v6(const std::string & ip, std::unordered_set<std::string> & response);
|
||||
|
||||
std::span<pollfd> get_readable_sockets(int * sockets, pollfd * pollfd);
|
||||
|
||||
int64_t calculate_timeout();
|
||||
|
||||
void process_possible_timeout();
|
||||
|
||||
void process_readable_sockets(std::span<pollfd> readable_sockets);
|
||||
|
||||
ares_channel channel;
|
||||
|
||||
static std::mutex mutex;
|
||||
};
|
||||
}
|
||||
|
||||
|
||||
@ -5,8 +5,10 @@ namespace DB
|
||||
{
|
||||
std::shared_ptr<DNSPTRResolver> DNSPTRResolverProvider::get()
|
||||
{
|
||||
return std::make_shared<CaresPTRResolver>(
|
||||
static auto resolver = std::make_shared<CaresPTRResolver>(
|
||||
CaresPTRResolver::provider_token {}
|
||||
);
|
||||
|
||||
return resolver;
|
||||
}
|
||||
}
|
||||
|
||||
@ -13,9 +13,9 @@ namespace DB
|
||||
|
||||
using Checksum = CityHash_v1_0_2::uint128;
|
||||
|
||||
CompressionCodecPtr getCompressionCodecForFile(const DataPartStoragePtr & data_part_storage, const String & relative_path)
|
||||
CompressionCodecPtr getCompressionCodecForFile(const IDataPartStorage & data_part_storage, const String & relative_path)
|
||||
{
|
||||
auto read_buffer = data_part_storage->readFile(relative_path, {}, std::nullopt, std::nullopt);
|
||||
auto read_buffer = data_part_storage.readFile(relative_path, {}, std::nullopt, std::nullopt);
|
||||
read_buffer->ignore(sizeof(Checksum));
|
||||
|
||||
UInt8 header_size = ICompressionCodec::getHeaderSize();
|
||||
|
||||
@ -11,6 +11,6 @@ namespace DB
|
||||
/// clickhouse fashion (with checksums, headers for each block, etc). This
|
||||
/// method should be used as fallback when we cannot deduce compression codec
|
||||
/// from metadata.
|
||||
CompressionCodecPtr getCompressionCodecForFile(const DataPartStoragePtr & data_part_storage, const String & relative_path);
|
||||
CompressionCodecPtr getCompressionCodecForFile(const IDataPartStorage & data_part_storage, const String & relative_path);
|
||||
|
||||
}
|
||||
|
||||
@ -1049,7 +1049,7 @@ INSTANTIATE_TEST_SUITE_P(RandomInt,
|
||||
::testing::Combine(
|
||||
DefaultCodecsToTest,
|
||||
::testing::Values(
|
||||
generateSeq<UInt8 >(G(RandomGenerator<UInt8>(0))),
|
||||
generateSeq<UInt8 >(G(RandomGenerator<uint8_t>(0))),
|
||||
generateSeq<UInt16>(G(RandomGenerator<UInt16>(0))),
|
||||
generateSeq<UInt32>(G(RandomGenerator<UInt32>(0, 0, 1000'000'000))),
|
||||
generateSeq<UInt64>(G(RandomGenerator<UInt64>(0, 0, 1000'000'000)))
|
||||
|
||||
@ -69,7 +69,7 @@ static std::ostream & operator<<(std::ostream & ostr, const JSONPathAndValue & p
|
||||
bool first = true;
|
||||
for (const auto & part : path_and_value.path.getParts())
|
||||
{
|
||||
ostr << (first ? "{" : ", {") << part.key << ", " << part.is_nested << ", " << part.anonymous_array_level << "}";
|
||||
ostr << (first ? "{" : ", {") << part.key << ", " << part.is_nested << ", " << static_cast<uint8_t>(part.anonymous_array_level) << "}";
|
||||
first = false;
|
||||
}
|
||||
|
||||
|
||||
@ -31,4 +31,33 @@ VolumePtr IStoragePolicy::getVolumeByName(const String & volume_name) const
|
||||
return volume;
|
||||
}
|
||||
|
||||
size_t IStoragePolicy::getVolumeIndexByDiskName(const String & disk_name) const
|
||||
{
|
||||
auto index = tryGetVolumeIndexByDiskName(disk_name);
|
||||
if (!index)
|
||||
throw Exception(ErrorCodes::UNKNOWN_DISK,
|
||||
"No disk {} in policy {}", backQuote(disk_name), backQuote(getName()));
|
||||
|
||||
return *index;
|
||||
}
|
||||
|
||||
VolumePtr IStoragePolicy::tryGetVolumeByDiskName(const String & disk_name) const
|
||||
{
|
||||
auto index = tryGetVolumeIndexByDiskName(disk_name);
|
||||
if (!index)
|
||||
return nullptr;
|
||||
|
||||
return getVolume(*index);
|
||||
}
|
||||
|
||||
VolumePtr IStoragePolicy::getVolumeByDiskName(const String & disk_name) const
|
||||
{
|
||||
auto volume = tryGetVolumeByDiskName(disk_name);
|
||||
if (!volume)
|
||||
throw Exception(ErrorCodes::UNKNOWN_DISK,
|
||||
"No disk {} in policy {}", backQuote(disk_name), backQuote(getName()));
|
||||
|
||||
return volume;
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -4,6 +4,7 @@
|
||||
|
||||
#include <memory>
|
||||
#include <vector>
|
||||
#include <optional>
|
||||
#include <base/types.h>
|
||||
|
||||
namespace DB
|
||||
@ -55,12 +56,15 @@ public:
|
||||
/// Get volume by index.
|
||||
virtual VolumePtr getVolume(size_t index) const = 0;
|
||||
virtual VolumePtr tryGetVolumeByName(const String & volume_name) const = 0;
|
||||
virtual VolumePtr tryGetVolumeByDisk(const DiskPtr & disk_ptr) const = 0;
|
||||
VolumePtr getVolumeByName(const String & volume_name) const;
|
||||
/// Checks if storage policy can be replaced by another one.
|
||||
virtual void checkCompatibleWith(const StoragePolicyPtr & new_storage_policy) const = 0;
|
||||
/// Find volume index, which contains disk
|
||||
virtual size_t getVolumeIndexByDisk(const DiskPtr & disk_ptr) const = 0;
|
||||
/// Finds a volume index, which contains disk
|
||||
virtual std::optional<size_t> tryGetVolumeIndexByDiskName(const String & disk_name) const = 0;
|
||||
size_t getVolumeIndexByDiskName(const String & disk_name) const;
|
||||
/// Finds a volume which contains a specified disk.
|
||||
VolumePtr tryGetVolumeByDiskName(const String & disk_name) const;
|
||||
VolumePtr getVolumeByDiskName(const String & disk_name) const;
|
||||
/// Check if we have any volume with stopped merges
|
||||
virtual bool hasAnyVolumeWithDisabledMerges() const = 0;
|
||||
virtual bool containsVolume(const String & volume_name) const = 0;
|
||||
|
||||
@ -599,7 +599,7 @@ std::unique_ptr<WriteBufferFromFileBase> DiskObjectStorageTransaction::writeFile
|
||||
auto write_operation = std::make_unique<WriteFileObjectStorageOperation>(object_storage, metadata_storage, object);
|
||||
std::function<void(size_t count)> create_metadata_callback;
|
||||
|
||||
if (autocommit)
|
||||
if (autocommit)
|
||||
{
|
||||
create_metadata_callback = [tx = shared_from_this(), mode, path, blob_name] (size_t count)
|
||||
{
|
||||
|
||||
@ -26,7 +26,6 @@ namespace ErrorCodes
|
||||
extern const int BAD_ARGUMENTS;
|
||||
extern const int EXCESSIVE_ELEMENT_IN_CONFIG;
|
||||
extern const int NO_ELEMENTS_IN_CONFIG;
|
||||
extern const int UNKNOWN_DISK;
|
||||
extern const int UNKNOWN_POLICY;
|
||||
extern const int UNKNOWN_VOLUME;
|
||||
extern const int LOGICAL_ERROR;
|
||||
@ -311,22 +310,12 @@ void StoragePolicy::checkCompatibleWith(const StoragePolicyPtr & new_storage_pol
|
||||
}
|
||||
|
||||
|
||||
size_t StoragePolicy::getVolumeIndexByDisk(const DiskPtr & disk_ptr) const
|
||||
std::optional<size_t> StoragePolicy::tryGetVolumeIndexByDiskName(const String & disk_name) const
|
||||
{
|
||||
auto it = volume_index_by_disk_name.find(disk_ptr->getName());
|
||||
auto it = volume_index_by_disk_name.find(disk_name);
|
||||
if (it != volume_index_by_disk_name.end())
|
||||
return it->second;
|
||||
else
|
||||
throw Exception("No disk " + backQuote(disk_ptr->getName()) + " in policy " + backQuote(name), ErrorCodes::UNKNOWN_DISK);
|
||||
}
|
||||
|
||||
|
||||
VolumePtr StoragePolicy::tryGetVolumeByDisk(const DiskPtr & disk_ptr) const
|
||||
{
|
||||
auto it = volume_index_by_disk_name.find(disk_ptr->getName());
|
||||
if (it == volume_index_by_disk_name.end())
|
||||
return nullptr;
|
||||
return getVolume(it->second);
|
||||
return {};
|
||||
}
|
||||
|
||||
|
||||
|
||||
@ -68,7 +68,7 @@ public:
|
||||
ReservationPtr reserve(UInt64 bytes, size_t min_volume_index) const override;
|
||||
|
||||
/// Find volume index, which contains disk
|
||||
size_t getVolumeIndexByDisk(const DiskPtr & disk_ptr) const override;
|
||||
std::optional<size_t> tryGetVolumeIndexByDiskName(const String & disk_name) const override;
|
||||
|
||||
/// Reserves 0 bytes on disk with max available space
|
||||
/// Do not use this function when it is possible to predict size.
|
||||
@ -85,9 +85,6 @@ public:
|
||||
|
||||
VolumePtr tryGetVolumeByName(const String & volume_name) const override;
|
||||
|
||||
/// Finds a volume which contains a specified disk.
|
||||
VolumePtr tryGetVolumeByDisk(const DiskPtr & disk_ptr) const override;
|
||||
|
||||
/// Checks if storage policy can be replaced by another one.
|
||||
void checkCompatibleWith(const StoragePolicyPtr & new_storage_policy) const override;
|
||||
|
||||
|
||||
@ -39,6 +39,7 @@
|
||||
#include <Common/FieldVisitorsAccurateComparison.h>
|
||||
#include <Common/assert_cast.h>
|
||||
#include <Common/typeid_cast.h>
|
||||
#include <DataTypes/DataTypeLowCardinality.h>
|
||||
#include <Interpreters/Context.h>
|
||||
|
||||
#if USE_EMBEDDED_COMPILER
|
||||
@ -1790,25 +1791,32 @@ public:
|
||||
// const +|- variable
|
||||
if (left.column && isColumnConst(*left.column))
|
||||
{
|
||||
auto left_type = removeNullable(removeLowCardinality(left.type));
|
||||
auto right_type = removeNullable(removeLowCardinality(right.type));
|
||||
auto ret_type = removeNullable(removeLowCardinality(return_type));
|
||||
|
||||
auto transform = [&](const Field & point)
|
||||
{
|
||||
ColumnsWithTypeAndName columns_with_constant
|
||||
= {{left.column->cloneResized(1), left.type, left.name},
|
||||
{right.type->createColumnConst(1, point), right.type, right.name}};
|
||||
= {{left_type->createColumnConst(1, (*left.column)[0]), left_type, left.name},
|
||||
{right_type->createColumnConst(1, point), right_type, right.name}};
|
||||
|
||||
auto col = Base::executeImpl(columns_with_constant, return_type, 1);
|
||||
/// This is a bit dangerous to call Base::executeImpl cause it ignores `use Default Implementation For XXX` flags.
|
||||
/// It was possible to check monotonicity for nullable right type which result to exception.
|
||||
/// Adding removeNullable above fixes the issue, but some other inconsistency may left.
|
||||
auto col = Base::executeImpl(columns_with_constant, ret_type, 1);
|
||||
Field point_transformed;
|
||||
col->get(0, point_transformed);
|
||||
return point_transformed;
|
||||
};
|
||||
transform(left_point);
|
||||
transform(right_point);
|
||||
|
||||
bool is_positive_monotonicity = applyVisitor(FieldVisitorAccurateLess(), left_point, right_point)
|
||||
== applyVisitor(FieldVisitorAccurateLess(), transform(left_point), transform(right_point));
|
||||
|
||||
if (name_view == "plus")
|
||||
{
|
||||
// Check if there is an overflow
|
||||
if (applyVisitor(FieldVisitorAccurateLess(), left_point, right_point)
|
||||
== applyVisitor(FieldVisitorAccurateLess(), transform(left_point), transform(right_point)))
|
||||
if (is_positive_monotonicity)
|
||||
return {true, true, false, true};
|
||||
else
|
||||
return {false, true, false, false};
|
||||
@ -1816,8 +1824,7 @@ public:
|
||||
else
|
||||
{
|
||||
// Check if there is an overflow
|
||||
if (applyVisitor(FieldVisitorAccurateLess(), left_point, right_point)
|
||||
!= applyVisitor(FieldVisitorAccurateLess(), transform(left_point), transform(right_point)))
|
||||
if (!is_positive_monotonicity)
|
||||
return {true, false, false, true};
|
||||
else
|
||||
return {false, false, false, false};
|
||||
@ -1826,13 +1833,17 @@ public:
|
||||
// variable +|- constant
|
||||
else if (right.column && isColumnConst(*right.column))
|
||||
{
|
||||
auto left_type = removeNullable(removeLowCardinality(left.type));
|
||||
auto right_type = removeNullable(removeLowCardinality(right.type));
|
||||
auto ret_type = removeNullable(removeLowCardinality(return_type));
|
||||
|
||||
auto transform = [&](const Field & point)
|
||||
{
|
||||
ColumnsWithTypeAndName columns_with_constant
|
||||
= {{left.type->createColumnConst(1, point), left.type, left.name},
|
||||
{right.column->cloneResized(1), right.type, right.name}};
|
||||
= {{left_type->createColumnConst(1, point), left_type, left.name},
|
||||
{right_type->createColumnConst(1, (*right.column)[0]), right_type, right.name}};
|
||||
|
||||
auto col = Base::executeImpl(columns_with_constant, return_type, 1);
|
||||
auto col = Base::executeImpl(columns_with_constant, ret_type, 1);
|
||||
Field point_transformed;
|
||||
col->get(0, point_transformed);
|
||||
return point_transformed;
|
||||
|
||||
@ -2173,6 +2173,10 @@ struct ToNumberMonotonicity
|
||||
const size_t size_of_from = type.getSizeOfValueInMemory();
|
||||
const size_t size_of_to = sizeof(T);
|
||||
|
||||
/// Do not support 128 bit integers and decimals for now.
|
||||
if (size_of_from > sizeof(Int64))
|
||||
return {};
|
||||
|
||||
const bool left_in_first_half = left.isNull()
|
||||
? from_is_unsigned
|
||||
: (left.get<Int64>() >= 0);
|
||||
|
||||
@ -43,10 +43,34 @@ using FunctionCutToFirstSignificantSubdomainWithWWWRFC = FunctionStringToString<
|
||||
|
||||
REGISTER_FUNCTION(CutToFirstSignificantSubdomain)
|
||||
{
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomain>();
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainWithWWW>();
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainRFC>();
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainWithWWWRFC>();
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomain>(
|
||||
{
|
||||
R"(Returns the part of the domain that includes top-level subdomains up to the "first significant subdomain" (see documentation of the `firstSignificantSubdomain`).)",
|
||||
Documentation::Examples{
|
||||
{"cutToFirstSignificantSubdomain1", "SELECT cutToFirstSignificantSubdomain('https://news.clickhouse.com.tr/')"},
|
||||
{"cutToFirstSignificantSubdomain2", "SELECT cutToFirstSignificantSubdomain('www.tr')"},
|
||||
{"cutToFirstSignificantSubdomain3", "SELECT cutToFirstSignificantSubdomain('tr')"},
|
||||
},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainWithWWW>(
|
||||
{
|
||||
R"(Returns the part of the domain that includes top-level subdomains up to the "first significant subdomain", without stripping "www".)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainRFC>(
|
||||
{
|
||||
R"(Similar to `cutToFirstSignificantSubdomain` but follows stricter rules to be compatible with RFC 3986 and less performant.)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainWithWWWRFC>(
|
||||
{
|
||||
R"(Similar to `cutToFirstSignificantSubdomainWithWWW` but follows stricter rules to be compatible with RFC 3986 and less performant.)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -42,10 +42,41 @@ using FunctionCutToFirstSignificantSubdomainCustomWithWWWRFC = FunctionCutToFirs
|
||||
|
||||
REGISTER_FUNCTION(CutToFirstSignificantSubdomainCustom)
|
||||
{
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainCustom>();
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainCustomWithWWW>();
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainCustomRFC>();
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainCustomWithWWWRFC>();
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainCustom>(
|
||||
{
|
||||
R"(
|
||||
Returns the part of the domain that includes top-level subdomains up to the first significant subdomain. Accepts custom TLD list name.
|
||||
|
||||
Can be useful if you need fresh TLD list or you have custom.
|
||||
)",
|
||||
Documentation::Examples{
|
||||
{"cutToFirstSignificantSubdomainCustom", "SELECT cutToFirstSignificantSubdomainCustom('bar.foo.there-is-no-such-domain', 'public_suffix_list');"},
|
||||
},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainCustomWithWWW>(
|
||||
{
|
||||
R"(
|
||||
Returns the part of the domain that includes top-level subdomains up to the first significant subdomain without stripping `www`.
|
||||
Accepts custom TLD list name from config.
|
||||
|
||||
Can be useful if you need fresh TLD list or you have custom.
|
||||
)",
|
||||
Documentation::Examples{{"cutToFirstSignificantSubdomainCustomWithWWW", "SELECT cutToFirstSignificantSubdomainCustomWithWWW('www.foo', 'public_suffix_list')"}},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainCustomRFC>(
|
||||
{
|
||||
R"(Similar to `cutToFirstSignificantSubdomainCustom` but follows stricter rules according to RFC 3986.)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
factory.registerFunction<FunctionCutToFirstSignificantSubdomainCustomWithWWWRFC>(
|
||||
{
|
||||
R"(Similar to `cutToFirstSignificantSubdomainCustomWithWWW` but follows stricter rules according to RFC 3986.)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -14,8 +14,24 @@ using FunctionDomainRFC = FunctionStringToString<ExtractSubstringImpl<ExtractDom
|
||||
|
||||
REGISTER_FUNCTION(Domain)
|
||||
{
|
||||
factory.registerFunction<FunctionDomain>();
|
||||
factory.registerFunction<FunctionDomainRFC>();
|
||||
factory.registerFunction<FunctionDomain>(
|
||||
{
|
||||
R"(
|
||||
Extracts the hostname from a URL.
|
||||
|
||||
The URL can be specified with or without a scheme.
|
||||
If the argument can't be parsed as URL, the function returns an empty string.
|
||||
)",
|
||||
Documentation::Examples{{"domain", "SELECT domain('svn+ssh://some.svn-hosting.com:80/repo/trunk')"}},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
|
||||
factory.registerFunction<FunctionDomainRFC>(
|
||||
{
|
||||
R"(Similar to `domain` but follows stricter rules to be compatible with RFC 3986 and less performant.)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -14,8 +14,23 @@ using FunctionDomainWithoutWWWRFC = FunctionStringToString<ExtractSubstringImpl<
|
||||
|
||||
REGISTER_FUNCTION(DomainWithoutWWW)
|
||||
{
|
||||
factory.registerFunction<FunctionDomainWithoutWWW>();
|
||||
factory.registerFunction<FunctionDomainWithoutWWWRFC>();
|
||||
factory.registerFunction<FunctionDomainWithoutWWW>(
|
||||
{
|
||||
R"(
|
||||
Extracts the hostname from a URL, removing the leading "www." if present.
|
||||
|
||||
The URL can be specified with or without a scheme.
|
||||
If the argument can't be parsed as URL, the function returns an empty string.
|
||||
)",
|
||||
Documentation::Examples{{"domainWithoutWWW", "SELECT domainWithoutWWW('https://www.clickhouse.com')"}},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
factory.registerFunction<FunctionDomainWithoutWWWRFC>(
|
||||
{
|
||||
R"(Similar to `domainWithoutWWW` but follows stricter rules to be compatible with RFC 3986 and less performant.)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -14,8 +14,28 @@ using FunctionFirstSignificantSubdomainRFC = FunctionStringToString<ExtractSubst
|
||||
|
||||
REGISTER_FUNCTION(FirstSignificantSubdomain)
|
||||
{
|
||||
factory.registerFunction<FunctionFirstSignificantSubdomain>();
|
||||
factory.registerFunction<FunctionFirstSignificantSubdomainRFC>();
|
||||
factory.registerFunction<FunctionFirstSignificantSubdomain>(
|
||||
{
|
||||
R"(
|
||||
Returns the "first significant subdomain".
|
||||
|
||||
The first significant subdomain is a second-level domain if it is 'com', 'net', 'org', or 'co'.
|
||||
Otherwise, it is a third-level domain.
|
||||
|
||||
For example, firstSignificantSubdomain('https://news.clickhouse.com/') = 'clickhouse', firstSignificantSubdomain ('https://news.clickhouse.com.tr/') = 'clickhouse'.
|
||||
|
||||
The list of "insignificant" second-level domains and other implementation details may change in the future.
|
||||
)",
|
||||
Documentation::Examples{{"firstSignificantSubdomain", "SELECT firstSignificantSubdomain('https://news.clickhouse.com/')"}},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
|
||||
factory.registerFunction<FunctionFirstSignificantSubdomainRFC>(
|
||||
{
|
||||
R"(Returns the "first significant subdomain" according to RFC 1034.)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -139,8 +139,18 @@ struct FunctionPortRFC : public FunctionPortImpl<true>
|
||||
|
||||
REGISTER_FUNCTION(Port)
|
||||
{
|
||||
factory.registerFunction<FunctionPort>();
|
||||
factory.registerFunction<FunctionPortRFC>();
|
||||
factory.registerFunction<FunctionPort>(
|
||||
{
|
||||
R"(Returns the port or `default_port` if there is no port in the URL (or in case of validation error).)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
factory.registerFunction<FunctionPortRFC>(
|
||||
{
|
||||
R"(Similar to `port`, but conforms to RFC 3986.)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -53,8 +53,23 @@ using FunctionTopLevelDomainRFC = FunctionStringToString<ExtractSubstringImpl<Ex
|
||||
|
||||
REGISTER_FUNCTION(TopLevelDomain)
|
||||
{
|
||||
factory.registerFunction<FunctionTopLevelDomain>();
|
||||
factory.registerFunction<FunctionTopLevelDomainRFC>();
|
||||
factory.registerFunction<FunctionTopLevelDomain>(
|
||||
{
|
||||
R"(
|
||||
Extracts the the top-level domain from a URL.
|
||||
|
||||
Returns an empty string if the argument cannot be parsed as a URL or does not contain a top-level domain.
|
||||
)",
|
||||
Documentation::Examples{{"topLevelDomain", "SELECT topLevelDomain('svn+ssh://www.some.svn-hosting.com:80/repo/trunk')"}},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
|
||||
factory.registerFunction<FunctionTopLevelDomainRFC>(
|
||||
{
|
||||
R"(Similar to topLevelDomain, but conforms to RFC 3986.)",
|
||||
Documentation::Examples{},
|
||||
Documentation::Categories{"URL"}
|
||||
});
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -213,7 +213,7 @@ BlockIO InterpreterDropQuery::executeToTableImpl(ContextPtr context_, ASTDropQue
|
||||
{
|
||||
/// And for simple MergeTree we can stop merges before acquiring the lock
|
||||
auto merges_blocker = table->getActionLock(ActionLocks::PartsMerge);
|
||||
auto table_lock = table->lockExclusively(context_->getCurrentQueryId(), context_->getSettingsRef().lock_acquire_timeout);
|
||||
table_lock = table->lockExclusively(context_->getCurrentQueryId(), context_->getSettingsRef().lock_acquire_timeout);
|
||||
}
|
||||
|
||||
auto metadata_snapshot = table->getInMemoryMetadataPtr();
|
||||
|
||||
@ -2146,6 +2146,8 @@ void InterpreterSelectQuery::executeFetchColumns(QueryProcessingStage::Enum proc
|
||||
|
||||
auto [limit_length, limit_offset] = getLimitLengthAndOffset(query, context);
|
||||
|
||||
auto local_limits = getStorageLimits(*context, options);
|
||||
|
||||
/** Optimization - if not specified DISTINCT, WHERE, GROUP, HAVING, ORDER, JOIN, LIMIT BY, WITH TIES
|
||||
* but LIMIT is specified, and limit + offset < max_block_size,
|
||||
* then as the block size we will use limit + offset (not to read more from the table than requested),
|
||||
@ -2164,17 +2166,22 @@ void InterpreterSelectQuery::executeFetchColumns(QueryProcessingStage::Enum proc
|
||||
&& !query_analyzer->hasAggregation()
|
||||
&& !query_analyzer->hasWindow()
|
||||
&& query.limitLength()
|
||||
&& limit_length <= std::numeric_limits<UInt64>::max() - limit_offset
|
||||
&& limit_length + limit_offset < max_block_size)
|
||||
&& limit_length <= std::numeric_limits<UInt64>::max() - limit_offset)
|
||||
{
|
||||
max_block_size = std::max<UInt64>(1, limit_length + limit_offset);
|
||||
max_threads_execute_query = max_streams = 1;
|
||||
if (limit_length + limit_offset < max_block_size)
|
||||
{
|
||||
max_block_size = std::max<UInt64>(1, limit_length + limit_offset);
|
||||
max_threads_execute_query = max_streams = 1;
|
||||
}
|
||||
if (limit_length + limit_offset < local_limits.local_limits.size_limits.max_rows)
|
||||
{
|
||||
query_info.limit = limit_length + limit_offset;
|
||||
}
|
||||
}
|
||||
|
||||
if (!max_block_size)
|
||||
throw Exception("Setting 'max_block_size' cannot be zero", ErrorCodes::PARAMETER_OUT_OF_BOUND);
|
||||
|
||||
auto local_limits = getStorageLimits(*context, options);
|
||||
storage_limits.emplace_back(local_limits);
|
||||
|
||||
/// Initialize the initial data streams to which the query transforms are superimposed. Table or subquery or prepared input?
|
||||
|
||||
@ -207,8 +207,8 @@ bool PartLog::addNewParts(
|
||||
elem.table_name = table_id.table_name;
|
||||
elem.partition_id = part->info.partition_id;
|
||||
elem.part_name = part->name;
|
||||
elem.disk_name = part->data_part_storage->getDiskName();
|
||||
elem.path_on_disk = part->data_part_storage->getFullPath();
|
||||
elem.disk_name = part->getDataPartStorage().getDiskName();
|
||||
elem.path_on_disk = part->getDataPartStorage().getFullPath();
|
||||
elem.part_type = part->getType();
|
||||
|
||||
elem.bytes_compressed_on_disk = part->getBytesOnDisk();
|
||||
|
||||
@ -131,6 +131,7 @@ void Set::setHeader(const ColumnsWithTypeAndName & header)
|
||||
if (const auto * low_cardinality_type = typeid_cast<const DataTypeLowCardinality *>(data_types.back().get()))
|
||||
{
|
||||
data_types.back() = low_cardinality_type->getDictionaryType();
|
||||
set_elements_types.back() = low_cardinality_type->getDictionaryType();
|
||||
materialized_columns.emplace_back(key_columns.back()->convertToFullColumnIfLowCardinality());
|
||||
key_columns.back() = materialized_columns.back().get();
|
||||
}
|
||||
|
||||
@ -1,7 +1,5 @@
|
||||
#include <Processors/Merges/Algorithms/AggregatingSortedAlgorithm.h>
|
||||
|
||||
#include <Columns/ColumnAggregateFunction.h>
|
||||
#include <Common/AlignedBuffer.h>
|
||||
#include <DataTypes/DataTypeAggregateFunction.h>
|
||||
#include <DataTypes/DataTypeCustomSimpleAggregateFunction.h>
|
||||
#include <DataTypes/DataTypeLowCardinality.h>
|
||||
@ -18,70 +16,6 @@ AggregatingSortedAlgorithm::ColumnsDefinition::ColumnsDefinition() = default;
|
||||
AggregatingSortedAlgorithm::ColumnsDefinition::ColumnsDefinition(ColumnsDefinition &&) noexcept = default;
|
||||
AggregatingSortedAlgorithm::ColumnsDefinition::~ColumnsDefinition() = default;
|
||||
|
||||
/// Stores information for aggregation of AggregateFunction columns
|
||||
struct AggregatingSortedAlgorithm::AggregateDescription
|
||||
{
|
||||
ColumnAggregateFunction * column = nullptr;
|
||||
const size_t column_number = 0; /// Position in header.
|
||||
|
||||
AggregateDescription() = default;
|
||||
explicit AggregateDescription(size_t col_number) : column_number(col_number) {}
|
||||
};
|
||||
|
||||
/// Stores information for aggregation of SimpleAggregateFunction columns
|
||||
struct AggregatingSortedAlgorithm::SimpleAggregateDescription
|
||||
{
|
||||
/// An aggregate function 'anyLast', 'sum'...
|
||||
AggregateFunctionPtr function;
|
||||
IAggregateFunction::AddFunc add_function = nullptr;
|
||||
|
||||
size_t column_number = 0;
|
||||
IColumn * column = nullptr;
|
||||
|
||||
/// For LowCardinality, convert is converted to nested type. nested_type is nullptr if no conversion needed.
|
||||
const DataTypePtr nested_type; /// Nested type for LowCardinality, if it is.
|
||||
const DataTypePtr real_type; /// Type in header.
|
||||
|
||||
AlignedBuffer state;
|
||||
bool created = false;
|
||||
|
||||
SimpleAggregateDescription(
|
||||
AggregateFunctionPtr function_, const size_t column_number_,
|
||||
DataTypePtr nested_type_, DataTypePtr real_type_)
|
||||
: function(std::move(function_)), column_number(column_number_)
|
||||
, nested_type(std::move(nested_type_)), real_type(std::move(real_type_))
|
||||
{
|
||||
add_function = function->getAddressOfAddFunction();
|
||||
state.reset(function->sizeOfData(), function->alignOfData());
|
||||
}
|
||||
|
||||
void createState()
|
||||
{
|
||||
if (created)
|
||||
return;
|
||||
function->create(state.data());
|
||||
created = true;
|
||||
}
|
||||
|
||||
void destroyState()
|
||||
{
|
||||
if (!created)
|
||||
return;
|
||||
function->destroy(state.data());
|
||||
created = false;
|
||||
}
|
||||
|
||||
/// Explicitly destroy aggregation state if the stream is terminated
|
||||
~SimpleAggregateDescription()
|
||||
{
|
||||
destroyState();
|
||||
}
|
||||
|
||||
SimpleAggregateDescription() = default;
|
||||
SimpleAggregateDescription(SimpleAggregateDescription &&) = default;
|
||||
SimpleAggregateDescription(const SimpleAggregateDescription &) = delete;
|
||||
};
|
||||
|
||||
static AggregatingSortedAlgorithm::ColumnsDefinition defineColumns(
|
||||
const Block & header, const SortDescription & description)
|
||||
{
|
||||
@ -191,6 +125,39 @@ static void postprocessChunk(Chunk & chunk, const AggregatingSortedAlgorithm::Co
|
||||
}
|
||||
|
||||
|
||||
AggregatingSortedAlgorithm::SimpleAggregateDescription::SimpleAggregateDescription(
|
||||
AggregateFunctionPtr function_, const size_t column_number_,
|
||||
DataTypePtr nested_type_, DataTypePtr real_type_)
|
||||
: function(std::move(function_)), column_number(column_number_)
|
||||
, nested_type(std::move(nested_type_)), real_type(std::move(real_type_))
|
||||
{
|
||||
add_function = function->getAddressOfAddFunction();
|
||||
state.reset(function->sizeOfData(), function->alignOfData());
|
||||
}
|
||||
|
||||
void AggregatingSortedAlgorithm::SimpleAggregateDescription::createState()
|
||||
{
|
||||
if (created)
|
||||
return;
|
||||
function->create(state.data());
|
||||
created = true;
|
||||
}
|
||||
|
||||
void AggregatingSortedAlgorithm::SimpleAggregateDescription::destroyState()
|
||||
{
|
||||
if (!created)
|
||||
return;
|
||||
function->destroy(state.data());
|
||||
created = false;
|
||||
}
|
||||
|
||||
/// Explicitly destroy aggregation state if the stream is terminated
|
||||
AggregatingSortedAlgorithm::SimpleAggregateDescription::~SimpleAggregateDescription()
|
||||
{
|
||||
destroyState();
|
||||
}
|
||||
|
||||
|
||||
AggregatingSortedAlgorithm::AggregatingMergedData::AggregatingMergedData(
|
||||
MutableColumns columns_, UInt64 max_block_size_, ColumnsDefinition & def_)
|
||||
: MergedData(std::move(columns_), false, max_block_size_), def(def_)
|
||||
|
||||
@ -1,5 +1,7 @@
|
||||
#pragma once
|
||||
|
||||
#include <Columns/ColumnAggregateFunction.h>
|
||||
#include <Common/AlignedBuffer.h>
|
||||
#include <Processors/Merges/Algorithms/IMergingAlgorithmWithDelayedChunk.h>
|
||||
#include <Processors/Merges/Algorithms/MergedData.h>
|
||||
|
||||
@ -23,8 +25,48 @@ public:
|
||||
void consume(Input & input, size_t source_num) override;
|
||||
Status merge() override;
|
||||
|
||||
struct SimpleAggregateDescription;
|
||||
struct AggregateDescription;
|
||||
/// Stores information for aggregation of SimpleAggregateFunction columns
|
||||
struct SimpleAggregateDescription
|
||||
{
|
||||
/// An aggregate function 'anyLast', 'sum'...
|
||||
AggregateFunctionPtr function;
|
||||
IAggregateFunction::AddFunc add_function = nullptr;
|
||||
|
||||
size_t column_number = 0;
|
||||
IColumn * column = nullptr;
|
||||
|
||||
/// For LowCardinality, convert is converted to nested type. nested_type is nullptr if no conversion needed.
|
||||
const DataTypePtr nested_type; /// Nested type for LowCardinality, if it is.
|
||||
const DataTypePtr real_type; /// Type in header.
|
||||
|
||||
AlignedBuffer state;
|
||||
bool created = false;
|
||||
|
||||
SimpleAggregateDescription(
|
||||
AggregateFunctionPtr function_, const size_t column_number_,
|
||||
DataTypePtr nested_type_, DataTypePtr real_type_);
|
||||
|
||||
void createState();
|
||||
|
||||
void destroyState();
|
||||
|
||||
/// Explicitly destroy aggregation state if the stream is terminated
|
||||
~SimpleAggregateDescription();
|
||||
|
||||
SimpleAggregateDescription() = default;
|
||||
SimpleAggregateDescription(SimpleAggregateDescription &&) = default;
|
||||
SimpleAggregateDescription(const SimpleAggregateDescription &) = delete;
|
||||
};
|
||||
|
||||
/// Stores information for aggregation of AggregateFunction columns
|
||||
struct AggregateDescription
|
||||
{
|
||||
ColumnAggregateFunction * column = nullptr;
|
||||
const size_t column_number = 0; /// Position in header.
|
||||
|
||||
AggregateDescription() = default;
|
||||
explicit AggregateDescription(size_t col_number) : column_number(col_number) {}
|
||||
};
|
||||
|
||||
/// This structure define columns into one of three types:
|
||||
/// * columns which are not aggregate functions and not needed to be aggregated
|
||||
|
||||
@ -23,10 +23,6 @@ namespace ErrorCodes
|
||||
extern const int CORRUPTED_DATA;
|
||||
}
|
||||
|
||||
SummingSortedAlgorithm::ColumnsDefinition::ColumnsDefinition() = default;
|
||||
SummingSortedAlgorithm::ColumnsDefinition::ColumnsDefinition(ColumnsDefinition &&) noexcept = default;
|
||||
SummingSortedAlgorithm::ColumnsDefinition::~ColumnsDefinition() = default;
|
||||
|
||||
/// Stores numbers of key-columns and value-columns.
|
||||
struct SummingSortedAlgorithm::MapDescription
|
||||
{
|
||||
@ -777,4 +773,8 @@ IMergingAlgorithm::Status SummingSortedAlgorithm::merge()
|
||||
return Status(merged_data.pull(), true);
|
||||
}
|
||||
|
||||
SummingSortedAlgorithm::ColumnsDefinition::ColumnsDefinition() = default;
|
||||
SummingSortedAlgorithm::ColumnsDefinition::ColumnsDefinition(ColumnsDefinition &&) noexcept = default;
|
||||
SummingSortedAlgorithm::ColumnsDefinition::~ColumnsDefinition() = default;
|
||||
|
||||
}
|
||||
|
||||
@ -173,6 +173,9 @@ Pipe ReadFromMergeTree::readFromPool(
|
||||
total_rows += part.getRowsCount();
|
||||
}
|
||||
|
||||
if (query_info.limit > 0 && query_info.limit < total_rows)
|
||||
total_rows = query_info.limit;
|
||||
|
||||
const auto & settings = context->getSettingsRef();
|
||||
const auto & client_info = context->getClientInfo();
|
||||
MergeTreeReadPool::BackoffSettings backoff_settings(settings);
|
||||
@ -246,10 +249,26 @@ ProcessorPtr ReadFromMergeTree::createSource(
|
||||
};
|
||||
}
|
||||
|
||||
return std::make_shared<TSource>(
|
||||
auto total_rows = part.getRowsCount();
|
||||
if (query_info.limit > 0 && query_info.limit < total_rows)
|
||||
total_rows = query_info.limit;
|
||||
|
||||
/// Actually it means that parallel reading from replicas enabled
|
||||
/// and we have to collaborate with initiator.
|
||||
/// In this case we won't set approximate rows, because it will be accounted multiple times.
|
||||
/// Also do not count amount of read rows if we read in order of sorting key,
|
||||
/// because we don't know actual amount of read rows in case when limit is set.
|
||||
bool set_rows_approx = !extension.has_value() && !reader_settings.read_in_order;
|
||||
|
||||
auto source = std::make_shared<TSource>(
|
||||
data, storage_snapshot, part.data_part, max_block_size, preferred_block_size_bytes,
|
||||
preferred_max_column_in_block_size_bytes, required_columns, part.ranges, use_uncompressed_cache, prewhere_info,
|
||||
actions_settings, reader_settings, virt_column_names, part.part_index_in_query, has_limit_below_one_block, std::move(extension));
|
||||
|
||||
if (set_rows_approx)
|
||||
source -> addTotalRowsApprox(total_rows);
|
||||
|
||||
return source;
|
||||
}
|
||||
|
||||
Pipe ReadFromMergeTree::readInOrder(
|
||||
|
||||
@ -6,12 +6,12 @@
|
||||
#include <IO/ReadBufferFromFileBase.h>
|
||||
#include <IO/ReadHelpers.h>
|
||||
#include <Common/logger_useful.h>
|
||||
#include <Disks/IStoragePolicy.h>
|
||||
#include <Backups/BackupEntryFromSmallFile.h>
|
||||
#include <Backups/BackupEntryFromImmutableFile.h>
|
||||
#include <Storages/MergeTree/localBackup.h>
|
||||
#include <Disks/SingleDiskVolume.h>
|
||||
#include <Interpreters/TransactionVersionMetadata.h>
|
||||
#include <memory>
|
||||
|
||||
namespace DB
|
||||
{
|
||||
@ -29,6 +29,16 @@ DataPartStorageOnDisk::DataPartStorageOnDisk(VolumePtr volume_, std::string root
|
||||
{
|
||||
}
|
||||
|
||||
DataPartStorageOnDisk::DataPartStorageOnDisk(
|
||||
VolumePtr volume_, std::string root_path_, std::string part_dir_, DiskTransactionPtr transaction_)
|
||||
: volume(std::move(volume_))
|
||||
, root_path(std::move(root_path_))
|
||||
, part_dir(std::move(part_dir_))
|
||||
, transaction(std::move(transaction_))
|
||||
, has_shared_transaction(transaction != nullptr)
|
||||
{
|
||||
}
|
||||
|
||||
std::string DataPartStorageOnDisk::getFullPath() const
|
||||
{
|
||||
return fs::path(volume->getDisk()->getPath()) / root_path / part_dir / "";
|
||||
@ -49,6 +59,11 @@ std::string DataPartStorageOnDisk::getFullRootPath() const
|
||||
return fs::path(volume->getDisk()->getPath()) / root_path / "";
|
||||
}
|
||||
|
||||
MutableDataPartStoragePtr DataPartStorageOnDisk::getProjection(const std::string & name)
|
||||
{
|
||||
return std::shared_ptr<DataPartStorageOnDisk>(new DataPartStorageOnDisk(volume, std::string(fs::path(root_path) / part_dir), name, transaction));
|
||||
}
|
||||
|
||||
DataPartStoragePtr DataPartStorageOnDisk::getProjection(const std::string & name) const
|
||||
{
|
||||
return std::make_shared<DataPartStorageOnDisk>(volume, std::string(fs::path(root_path) / part_dir), name);
|
||||
@ -113,6 +128,7 @@ static UInt64 calculateTotalSizeOnDiskImpl(const DiskPtr & disk, const String &
|
||||
{
|
||||
if (disk->isFile(from))
|
||||
return disk->getFileSize(from);
|
||||
|
||||
std::vector<std::string> files;
|
||||
disk->listFiles(from, files);
|
||||
UInt64 res = 0;
|
||||
@ -135,75 +151,11 @@ std::unique_ptr<ReadBufferFromFileBase> DataPartStorageOnDisk::readFile(
|
||||
return volume->getDisk()->readFile(fs::path(root_path) / part_dir / name, settings, read_hint, file_size);
|
||||
}
|
||||
|
||||
static std::unique_ptr<ReadBufferFromFileBase> openForReading(const DiskPtr & disk, const String & path)
|
||||
{
|
||||
size_t file_size = disk->getFileSize(path);
|
||||
return disk->readFile(path, ReadSettings().adjustBufferSize(file_size), file_size);
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::loadVersionMetadata(VersionMetadata & version, Poco::Logger * log) const
|
||||
{
|
||||
std::string version_file_name = fs::path(root_path) / part_dir / "txn_version.txt";
|
||||
String tmp_version_file_name = version_file_name + ".tmp";
|
||||
DiskPtr disk = volume->getDisk();
|
||||
|
||||
auto remove_tmp_file = [&]()
|
||||
{
|
||||
auto last_modified = disk->getLastModified(tmp_version_file_name);
|
||||
auto buf = openForReading(disk, tmp_version_file_name);
|
||||
String content;
|
||||
readStringUntilEOF(content, *buf);
|
||||
LOG_WARNING(log, "Found file {} that was last modified on {}, has size {} and the following content: {}",
|
||||
tmp_version_file_name, last_modified.epochTime(), content.size(), content);
|
||||
disk->removeFile(tmp_version_file_name);
|
||||
};
|
||||
|
||||
if (disk->exists(version_file_name))
|
||||
{
|
||||
auto buf = openForReading(disk, version_file_name);
|
||||
version.read(*buf);
|
||||
if (disk->exists(tmp_version_file_name))
|
||||
remove_tmp_file();
|
||||
return;
|
||||
}
|
||||
|
||||
/// Four (?) cases are possible:
|
||||
/// 1. Part was created without transactions.
|
||||
/// 2. Version metadata file was not renamed from *.tmp on part creation.
|
||||
/// 3. Version metadata were written to *.tmp file, but hard restart happened before fsync.
|
||||
/// 4. Fsyncs in storeVersionMetadata() work incorrectly.
|
||||
|
||||
if (!disk->exists(tmp_version_file_name))
|
||||
{
|
||||
/// Case 1.
|
||||
/// We do not have version metadata and transactions history for old parts,
|
||||
/// so let's consider that such parts were created by some ancient transaction
|
||||
/// and were committed with some prehistoric CSN.
|
||||
/// NOTE It might be Case 3, but version metadata file is written on part creation before other files,
|
||||
/// so it's not Case 3 if part is not broken.
|
||||
version.setCreationTID(Tx::PrehistoricTID, nullptr);
|
||||
version.creation_csn = Tx::PrehistoricCSN;
|
||||
return;
|
||||
}
|
||||
|
||||
/// Case 2.
|
||||
/// Content of *.tmp file may be broken, just use fake TID.
|
||||
/// Transaction was not committed if *.tmp file was not renamed, so we should complete rollback by removing part.
|
||||
version.setCreationTID(Tx::DummyTID, nullptr);
|
||||
version.creation_csn = Tx::RolledBackCSN;
|
||||
remove_tmp_file();
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::checkConsistency(const MergeTreeDataPartChecksums & checksums) const
|
||||
{
|
||||
checksums.checkSizes(volume->getDisk(), getRelativePath());
|
||||
}
|
||||
|
||||
DataPartStorageBuilderPtr DataPartStorageOnDisk::getBuilder() const
|
||||
{
|
||||
return std::make_shared<DataPartStorageBuilderOnDisk>(volume, root_path, part_dir);
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::remove(
|
||||
CanRemoveCallback && can_remove_callback,
|
||||
const MergeTreeDataPartChecksums & checksums,
|
||||
@ -273,7 +225,7 @@ void DataPartStorageOnDisk::remove(
|
||||
try
|
||||
{
|
||||
disk->moveDirectory(from, to);
|
||||
onRename(root_path, part_dir_without_slash);
|
||||
part_dir = part_dir_without_slash;
|
||||
}
|
||||
catch (const Exception & e)
|
||||
{
|
||||
@ -488,11 +440,6 @@ bool DataPartStorageOnDisk::looksLikeBrokenDetachedPartHasTheSameContent(const S
|
||||
return original_files_list == detached_files_list;
|
||||
}
|
||||
|
||||
void DataPartStorageBuilderOnDisk::setRelativePath(const std::string & path)
|
||||
{
|
||||
part_dir = path;
|
||||
}
|
||||
|
||||
std::string DataPartStorageOnDisk::getDiskName() const
|
||||
{
|
||||
return volume->getDisk()->getName();
|
||||
@ -523,7 +470,7 @@ bool DataPartStorageOnDisk::isBroken() const
|
||||
return volume->getDisk()->isBroken();
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::syncRevision(UInt64 revision)
|
||||
void DataPartStorageOnDisk::syncRevision(UInt64 revision) const
|
||||
{
|
||||
volume->getDisk()->syncRevision(revision);
|
||||
}
|
||||
@ -543,11 +490,6 @@ std::string DataPartStorageOnDisk::getDiskPath() const
|
||||
return volume->getDisk()->getPath();
|
||||
}
|
||||
|
||||
DataPartStorageOnDisk::DisksSet::const_iterator DataPartStorageOnDisk::isStoredOnDisk(const DisksSet & disks) const
|
||||
{
|
||||
return disks.find(volume->getDisk());
|
||||
}
|
||||
|
||||
ReservationPtr DataPartStorageOnDisk::reserve(UInt64 bytes) const
|
||||
{
|
||||
auto res = volume->reserve(bytes);
|
||||
@ -562,159 +504,6 @@ ReservationPtr DataPartStorageOnDisk::tryReserve(UInt64 bytes) const
|
||||
return volume->reserve(bytes);
|
||||
}
|
||||
|
||||
size_t DataPartStorageOnDisk::getVolumeIndex(const IStoragePolicy & storage_policy) const
|
||||
{
|
||||
return storage_policy.getVolumeIndexByDisk(volume->getDisk());
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::writeChecksums(const MergeTreeDataPartChecksums & checksums, const WriteSettings & settings) const
|
||||
{
|
||||
std::string path = fs::path(root_path) / part_dir / "checksums.txt";
|
||||
|
||||
try
|
||||
{
|
||||
{
|
||||
auto out = volume->getDisk()->writeFile(path + ".tmp", 4096, WriteMode::Rewrite, settings);
|
||||
checksums.write(*out);
|
||||
}
|
||||
|
||||
volume->getDisk()->moveFile(path + ".tmp", path);
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
try
|
||||
{
|
||||
if (volume->getDisk()->exists(path + ".tmp"))
|
||||
volume->getDisk()->removeFile(path + ".tmp");
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
tryLogCurrentException("DataPartStorageOnDisk");
|
||||
}
|
||||
|
||||
throw;
|
||||
}
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::writeColumns(const NamesAndTypesList & columns, const WriteSettings & settings) const
|
||||
{
|
||||
std::string path = fs::path(root_path) / part_dir / "columns.txt";
|
||||
|
||||
try
|
||||
{
|
||||
auto buf = volume->getDisk()->writeFile(path + ".tmp", 4096, WriteMode::Rewrite, settings);
|
||||
columns.writeText(*buf);
|
||||
buf->finalize();
|
||||
|
||||
volume->getDisk()->moveFile(path + ".tmp", path);
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
try
|
||||
{
|
||||
if (volume->getDisk()->exists(path + ".tmp"))
|
||||
volume->getDisk()->removeFile(path + ".tmp");
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
tryLogCurrentException("DataPartStorageOnDisk");
|
||||
}
|
||||
|
||||
throw;
|
||||
}
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::writeVersionMetadata(const VersionMetadata & version, bool fsync_part_dir) const
|
||||
{
|
||||
std::string path = fs::path(root_path) / part_dir / "txn_version.txt";
|
||||
try
|
||||
{
|
||||
{
|
||||
/// TODO IDisk interface does not allow to open file with O_EXCL flag (for DiskLocal),
|
||||
/// so we create empty file at first (expecting that createFile throws if file already exists)
|
||||
/// and then overwrite it.
|
||||
volume->getDisk()->createFile(path + ".tmp");
|
||||
auto buf = volume->getDisk()->writeFile(path + ".tmp", 256);
|
||||
version.write(*buf);
|
||||
buf->finalize();
|
||||
buf->sync();
|
||||
}
|
||||
|
||||
SyncGuardPtr sync_guard;
|
||||
if (fsync_part_dir)
|
||||
sync_guard = volume->getDisk()->getDirectorySyncGuard(getRelativePath());
|
||||
volume->getDisk()->replaceFile(path + ".tmp", path);
|
||||
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
try
|
||||
{
|
||||
if (volume->getDisk()->exists(path + ".tmp"))
|
||||
volume->getDisk()->removeFile(path + ".tmp");
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
tryLogCurrentException("DataPartStorageOnDisk");
|
||||
}
|
||||
|
||||
throw;
|
||||
}
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::appendCSNToVersionMetadata(const VersionMetadata & version, VersionMetadata::WhichCSN which_csn) const
|
||||
{
|
||||
/// Small enough appends to file are usually atomic,
|
||||
/// so we append new metadata instead of rewriting file to reduce number of fsyncs.
|
||||
/// We don't need to do fsync when writing CSN, because in case of hard restart
|
||||
/// we will be able to restore CSN from transaction log in Keeper.
|
||||
|
||||
std::string version_file_name = fs::path(root_path) / part_dir / "txn_version.txt";
|
||||
DiskPtr disk = volume->getDisk();
|
||||
auto out = disk->writeFile(version_file_name, 256, WriteMode::Append);
|
||||
version.writeCSN(*out, which_csn);
|
||||
out->finalize();
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::appendRemovalTIDToVersionMetadata(const VersionMetadata & version, bool clear) const
|
||||
{
|
||||
String version_file_name = fs::path(root_path) / part_dir / "txn_version.txt";
|
||||
DiskPtr disk = volume->getDisk();
|
||||
auto out = disk->writeFile(version_file_name, 256, WriteMode::Append);
|
||||
version.writeRemovalTID(*out, clear);
|
||||
out->finalize();
|
||||
|
||||
/// fsync is not required when we clearing removal TID, because after hard restart we will fix metadata
|
||||
if (!clear)
|
||||
out->sync();
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::writeDeleteOnDestroyMarker(Poco::Logger * log) const
|
||||
{
|
||||
String marker_path = fs::path(root_path) / part_dir / "delete-on-destroy.txt";
|
||||
auto disk = volume->getDisk();
|
||||
try
|
||||
{
|
||||
volume->getDisk()->createFile(marker_path);
|
||||
}
|
||||
catch (Poco::Exception & e)
|
||||
{
|
||||
LOG_ERROR(log, "{} (while creating DeleteOnDestroy marker: {})", e.what(), backQuote(fullPath(disk, marker_path)));
|
||||
}
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::removeDeleteOnDestroyMarker() const
|
||||
{
|
||||
std::string delete_on_destroy_file_name = fs::path(root_path) / part_dir / "delete-on-destroy.txt";
|
||||
volume->getDisk()->removeFileIfExists(delete_on_destroy_file_name);
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::removeVersionMetadata() const
|
||||
{
|
||||
std::string version_file_name = fs::path(root_path) / part_dir / "txn_version.txt";
|
||||
volume->getDisk()->removeFileIfExists(version_file_name);
|
||||
}
|
||||
|
||||
String DataPartStorageOnDisk::getUniqueId() const
|
||||
{
|
||||
auto disk = volume->getDisk();
|
||||
@ -724,16 +513,6 @@ String DataPartStorageOnDisk::getUniqueId() const
|
||||
return disk->getUniqueId(fs::path(getRelativePath()) / "checksums.txt");
|
||||
}
|
||||
|
||||
bool DataPartStorageOnDisk::shallParticipateInMerges(const IStoragePolicy & storage_policy) const
|
||||
{
|
||||
/// `IMergeTreeDataPart::volume` describes space where current part belongs, and holds
|
||||
/// `SingleDiskVolume` object which does not contain up-to-date settings of corresponding volume.
|
||||
/// Therefore we shall obtain volume from storage policy.
|
||||
auto volume_ptr = storage_policy.getVolume(storage_policy.getVolumeIndexByDisk(volume->getDisk()));
|
||||
|
||||
return !volume_ptr->areMergesAvoided();
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::backup(
|
||||
const MergeTreeDataPartChecksums & checksums,
|
||||
const NameSet & files_without_checksums,
|
||||
@ -798,7 +577,7 @@ void DataPartStorageOnDisk::backup(
|
||||
}
|
||||
}
|
||||
|
||||
DataPartStoragePtr DataPartStorageOnDisk::freeze(
|
||||
MutableDataPartStoragePtr DataPartStorageOnDisk::freeze(
|
||||
const std::string & to,
|
||||
const std::string & dir_path,
|
||||
bool make_source_readonly,
|
||||
@ -822,7 +601,7 @@ DataPartStoragePtr DataPartStorageOnDisk::freeze(
|
||||
return std::make_shared<DataPartStorageOnDisk>(single_disk_volume, to, dir_path);
|
||||
}
|
||||
|
||||
DataPartStoragePtr DataPartStorageOnDisk::clone(
|
||||
MutableDataPartStoragePtr DataPartStorageOnDisk::clonePart(
|
||||
const std::string & to,
|
||||
const std::string & dir_path,
|
||||
const DiskPtr & disk,
|
||||
@ -835,6 +614,7 @@ DataPartStoragePtr DataPartStorageOnDisk::clone(
|
||||
LOG_WARNING(log, "Path {} already exists. Will remove it and clone again.", fullPath(disk, path_to_clone));
|
||||
disk->removeRecursive(path_to_clone);
|
||||
}
|
||||
|
||||
disk->createDirectories(to);
|
||||
volume->getDisk()->copy(getRelativePath(), disk, to);
|
||||
volume->getDisk()->removeFileIfExists(fs::path(path_to_clone) / "delete-on-destroy.txt");
|
||||
@ -843,13 +623,7 @@ DataPartStoragePtr DataPartStorageOnDisk::clone(
|
||||
return std::make_shared<DataPartStorageOnDisk>(single_disk_volume, to, dir_path);
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::onRename(const std::string & new_root_path, const std::string & new_part_dir)
|
||||
{
|
||||
part_dir = new_part_dir;
|
||||
root_path = new_root_path;
|
||||
}
|
||||
|
||||
void DataPartStorageBuilderOnDisk::rename(
|
||||
void DataPartStorageOnDisk::rename(
|
||||
const std::string & new_root_path,
|
||||
const std::string & new_part_dir,
|
||||
Poco::Logger * log,
|
||||
@ -870,7 +644,7 @@ void DataPartStorageBuilderOnDisk::rename(
|
||||
"Part directory {} already exists and contains {} files. Removing it.",
|
||||
fullPath(volume->getDisk(), to), files.size());
|
||||
|
||||
transaction->removeRecursive(to);
|
||||
executeOperation([&](auto & disk) { disk.removeRecursive(to); });
|
||||
}
|
||||
else
|
||||
{
|
||||
@ -884,8 +658,12 @@ void DataPartStorageBuilderOnDisk::rename(
|
||||
String from = getRelativePath();
|
||||
|
||||
/// Why?
|
||||
transaction->setLastModified(from, Poco::Timestamp::fromEpochTime(time(nullptr)));
|
||||
transaction->moveDirectory(from, to);
|
||||
executeOperation([&](auto & disk)
|
||||
{
|
||||
disk.setLastModified(from, Poco::Timestamp::fromEpochTime(time(nullptr)));
|
||||
disk.moveDirectory(from, to);
|
||||
});
|
||||
|
||||
part_dir = new_part_dir;
|
||||
root_path = new_root_path;
|
||||
|
||||
@ -907,7 +685,7 @@ void DataPartStorageOnDisk::changeRootPath(const std::string & from_root, const
|
||||
--prefix_size;
|
||||
|
||||
if (prefix_size > root_path.size()
|
||||
|| std::string_view(from_root).substr(0, prefix_size) != std::string_view(root_path).substr(0, prefix_size))
|
||||
|| std::string_view(from_root).substr(0, prefix_size) != std::string_view(root_path).substr(0, prefix_size))
|
||||
throw Exception(
|
||||
ErrorCodes::LOGICAL_ERROR,
|
||||
"Cannot change part root to {} because it is not a prefix of current root {}",
|
||||
@ -920,51 +698,80 @@ void DataPartStorageOnDisk::changeRootPath(const std::string & from_root, const
|
||||
root_path = to_root.substr(0, dst_size) + root_path.substr(prefix_size);
|
||||
}
|
||||
|
||||
DataPartStorageBuilderOnDisk::DataPartStorageBuilderOnDisk(
|
||||
VolumePtr volume_,
|
||||
std::string root_path_,
|
||||
std::string part_dir_)
|
||||
: volume(std::move(volume_))
|
||||
, root_path(std::move(root_path_))
|
||||
, part_dir(std::move(part_dir_))
|
||||
, transaction(volume->getDisk()->createTransaction())
|
||||
{
|
||||
}
|
||||
|
||||
std::unique_ptr<WriteBufferFromFileBase> DataPartStorageBuilderOnDisk::writeFile(
|
||||
const String & name,
|
||||
size_t buf_size,
|
||||
const WriteSettings & settings)
|
||||
{
|
||||
return transaction->writeFile(fs::path(root_path) / part_dir / name, buf_size, WriteMode::Rewrite, settings, /* autocommit = */ false);
|
||||
}
|
||||
|
||||
void DataPartStorageBuilderOnDisk::removeFile(const String & name)
|
||||
{
|
||||
transaction->removeFile(fs::path(root_path) / part_dir / name);
|
||||
}
|
||||
|
||||
void DataPartStorageBuilderOnDisk::removeFileIfExists(const String & name)
|
||||
{
|
||||
transaction->removeFileIfExists(fs::path(root_path) / part_dir / name);
|
||||
}
|
||||
|
||||
void DataPartStorageBuilderOnDisk::removeRecursive()
|
||||
{
|
||||
transaction->removeRecursive(fs::path(root_path) / part_dir);
|
||||
}
|
||||
|
||||
void DataPartStorageBuilderOnDisk::removeSharedRecursive(bool keep_in_remote_fs)
|
||||
{
|
||||
transaction->removeSharedRecursive(fs::path(root_path) / part_dir, keep_in_remote_fs, {});
|
||||
}
|
||||
|
||||
SyncGuardPtr DataPartStorageBuilderOnDisk::getDirectorySyncGuard() const
|
||||
SyncGuardPtr DataPartStorageOnDisk::getDirectorySyncGuard() const
|
||||
{
|
||||
return volume->getDisk()->getDirectorySyncGuard(fs::path(root_path) / part_dir);
|
||||
}
|
||||
|
||||
void DataPartStorageBuilderOnDisk::createHardLinkFrom(const IDataPartStorage & source, const std::string & from, const std::string & to) const
|
||||
template <typename Op>
|
||||
void DataPartStorageOnDisk::executeOperation(Op && op)
|
||||
{
|
||||
if (transaction)
|
||||
op(*transaction);
|
||||
else
|
||||
op(*volume->getDisk());
|
||||
}
|
||||
|
||||
std::unique_ptr<WriteBufferFromFileBase> DataPartStorageOnDisk::writeFile(
|
||||
const String & name,
|
||||
size_t buf_size,
|
||||
const WriteSettings & settings)
|
||||
{
|
||||
if (transaction)
|
||||
return transaction->writeFile(fs::path(root_path) / part_dir / name, buf_size, WriteMode::Rewrite, settings, /* autocommit = */ false);
|
||||
|
||||
return volume->getDisk()->writeFile(fs::path(root_path) / part_dir / name, buf_size, WriteMode::Rewrite, settings);
|
||||
}
|
||||
|
||||
std::unique_ptr<WriteBufferFromFileBase> DataPartStorageOnDisk::writeTransactionFile(WriteMode mode) const
|
||||
{
|
||||
return volume->getDisk()->writeFile(fs::path(root_path) / part_dir / "txn_version.txt", 256, mode);
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::createFile(const String & name)
|
||||
{
|
||||
executeOperation([&](auto & disk) { disk.createFile(fs::path(root_path) / part_dir / name); });
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::moveFile(const String & from_name, const String & to_name)
|
||||
{
|
||||
executeOperation([&](auto & disk)
|
||||
{
|
||||
auto relative_path = fs::path(root_path) / part_dir;
|
||||
disk.moveFile(relative_path / from_name, relative_path / to_name);
|
||||
});
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::replaceFile(const String & from_name, const String & to_name)
|
||||
{
|
||||
executeOperation([&](auto & disk)
|
||||
{
|
||||
auto relative_path = fs::path(root_path) / part_dir;
|
||||
disk.replaceFile(relative_path / from_name, relative_path / to_name);
|
||||
});
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::removeFile(const String & name)
|
||||
{
|
||||
executeOperation([&](auto & disk) { disk.removeFile(fs::path(root_path) / part_dir / name); });
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::removeFileIfExists(const String & name)
|
||||
{
|
||||
executeOperation([&](auto & disk) { disk.removeFileIfExists(fs::path(root_path) / part_dir / name); });
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::removeRecursive()
|
||||
{
|
||||
executeOperation([&](auto & disk) { disk.removeRecursive(fs::path(root_path) / part_dir); });
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::removeSharedRecursive(bool keep_in_remote_fs)
|
||||
{
|
||||
executeOperation([&](auto & disk) { disk.removeSharedRecursive(fs::path(root_path) / part_dir, keep_in_remote_fs, {}); });
|
||||
}
|
||||
|
||||
void DataPartStorageOnDisk::createHardLinkFrom(const IDataPartStorage & source, const std::string & from, const std::string & to)
|
||||
{
|
||||
const auto * source_on_disk = typeid_cast<const DataPartStorageOnDisk *>(&source);
|
||||
if (!source_on_disk)
|
||||
@ -973,58 +780,43 @@ void DataPartStorageBuilderOnDisk::createHardLinkFrom(const IDataPartStorage & s
|
||||
"Cannot create hardlink from different storage. Expected DataPartStorageOnDisk, got {}",
|
||||
typeid(source).name());
|
||||
|
||||
transaction->createHardLink(
|
||||
fs::path(source_on_disk->getRelativePath()) / from,
|
||||
fs::path(root_path) / part_dir / to);
|
||||
executeOperation([&](auto & disk)
|
||||
{
|
||||
disk.createHardLink(
|
||||
fs::path(source_on_disk->getRelativePath()) / from,
|
||||
fs::path(root_path) / part_dir / to);
|
||||
});
|
||||
}
|
||||
|
||||
bool DataPartStorageBuilderOnDisk::exists() const
|
||||
void DataPartStorageOnDisk::createDirectories()
|
||||
{
|
||||
return volume->getDisk()->exists(fs::path(root_path) / part_dir);
|
||||
executeOperation([&](auto & disk) { disk.createDirectories(fs::path(root_path) / part_dir); });
|
||||
}
|
||||
|
||||
std::string DataPartStorageBuilderOnDisk::getFullPath() const
|
||||
void DataPartStorageOnDisk::createProjection(const std::string & name)
|
||||
{
|
||||
return fs::path(volume->getDisk()->getPath()) / root_path / part_dir;
|
||||
executeOperation([&](auto & disk) { disk.createDirectory(fs::path(root_path) / part_dir / name); });
|
||||
}
|
||||
|
||||
std::string DataPartStorageBuilderOnDisk::getRelativePath() const
|
||||
void DataPartStorageOnDisk::beginTransaction()
|
||||
{
|
||||
return fs::path(root_path) / part_dir;
|
||||
if (transaction)
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR,
|
||||
"Uncommitted {}transaction already exists", has_shared_transaction ? "shared " : "");
|
||||
|
||||
transaction = volume->getDisk()->createTransaction();
|
||||
}
|
||||
|
||||
void DataPartStorageBuilderOnDisk::createDirectories()
|
||||
void DataPartStorageOnDisk::commitTransaction()
|
||||
{
|
||||
transaction->createDirectories(fs::path(root_path) / part_dir);
|
||||
}
|
||||
if (!transaction)
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR, "There is no uncommitted transaction");
|
||||
|
||||
void DataPartStorageBuilderOnDisk::createProjection(const std::string & name)
|
||||
{
|
||||
transaction->createDirectory(fs::path(root_path) / part_dir / name);
|
||||
}
|
||||
if (has_shared_transaction)
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR, "Cannot commit shared transaction");
|
||||
|
||||
ReservationPtr DataPartStorageBuilderOnDisk::reserve(UInt64 bytes)
|
||||
{
|
||||
auto res = volume->reserve(bytes);
|
||||
if (!res)
|
||||
throw Exception(ErrorCodes::NOT_ENOUGH_SPACE, "Cannot reserve {}, not enough space", ReadableSize(bytes));
|
||||
|
||||
return res;
|
||||
}
|
||||
|
||||
DataPartStorageBuilderPtr DataPartStorageBuilderOnDisk::getProjection(const std::string & name) const
|
||||
{
|
||||
return std::make_shared<DataPartStorageBuilderOnDisk>(volume, std::string(fs::path(root_path) / part_dir), name);
|
||||
}
|
||||
|
||||
DataPartStoragePtr DataPartStorageBuilderOnDisk::getStorage() const
|
||||
{
|
||||
return std::make_shared<DataPartStorageOnDisk>(volume, root_path, part_dir);
|
||||
}
|
||||
|
||||
void DataPartStorageBuilderOnDisk::commit()
|
||||
{
|
||||
transaction->commit();
|
||||
transaction.reset();
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
@ -21,6 +21,7 @@ public:
|
||||
std::string getPartDirectory() const override { return part_dir; }
|
||||
std::string getFullRootPath() const override;
|
||||
|
||||
MutableDataPartStoragePtr getProjection(const std::string & name) override;
|
||||
DataPartStoragePtr getProjection(const std::string & name) const override;
|
||||
|
||||
bool exists() const override;
|
||||
@ -41,7 +42,6 @@ public:
|
||||
std::optional<size_t> read_hint,
|
||||
std::optional<size_t> file_size) const override;
|
||||
|
||||
void loadVersionMetadata(VersionMetadata & version, Poco::Logger * log) const override;
|
||||
void checkConsistency(const MergeTreeDataPartChecksums & checksums) const override;
|
||||
|
||||
void remove(
|
||||
@ -60,7 +60,6 @@ public:
|
||||
std::optional<Strings> & original_files_list) const;
|
||||
|
||||
void setRelativePath(const std::string & path) override;
|
||||
void onRename(const std::string & new_root_path, const std::string & new_part_dir) override;
|
||||
|
||||
std::string getDiskName() const override;
|
||||
std::string getDiskType() const override;
|
||||
@ -68,30 +67,14 @@ public:
|
||||
bool supportZeroCopyReplication() const override;
|
||||
bool supportParallelWrite() const override;
|
||||
bool isBroken() const override;
|
||||
void syncRevision(UInt64 revision) override;
|
||||
void syncRevision(UInt64 revision) const override;
|
||||
UInt64 getRevision() const override;
|
||||
std::unordered_map<String, String> getSerializedMetadata(const std::vector<String> & paths) const override;
|
||||
std::string getDiskPath() const override;
|
||||
|
||||
DisksSet::const_iterator isStoredOnDisk(const DisksSet & disks) const override;
|
||||
|
||||
ReservationPtr reserve(UInt64 bytes) const override;
|
||||
ReservationPtr tryReserve(UInt64 bytes) const override;
|
||||
size_t getVolumeIndex(const IStoragePolicy &) const override;
|
||||
|
||||
void writeChecksums(const MergeTreeDataPartChecksums & checksums, const WriteSettings & settings) const override;
|
||||
void writeColumns(const NamesAndTypesList & columns, const WriteSettings & settings) const override;
|
||||
void writeVersionMetadata(const VersionMetadata & version, bool fsync_part_dir) const override;
|
||||
void appendCSNToVersionMetadata(const VersionMetadata & version, VersionMetadata::WhichCSN which_csn) const override;
|
||||
void appendRemovalTIDToVersionMetadata(const VersionMetadata & version, bool clear) const override;
|
||||
void writeDeleteOnDestroyMarker(Poco::Logger * log) const override;
|
||||
void removeDeleteOnDestroyMarker() const override;
|
||||
void removeVersionMetadata() const override;
|
||||
|
||||
String getUniqueId() const override;
|
||||
|
||||
bool shallParticipateInMerges(const IStoragePolicy &) const override;
|
||||
|
||||
void backup(
|
||||
const MergeTreeDataPartChecksums & checksums,
|
||||
const NameSet & files_without_checksums,
|
||||
@ -100,7 +83,7 @@ public:
|
||||
bool make_temporary_hard_links,
|
||||
TemporaryFilesOnDisks * temp_dirs) const override;
|
||||
|
||||
DataPartStoragePtr freeze(
|
||||
MutableDataPartStoragePtr freeze(
|
||||
const std::string & to,
|
||||
const std::string & dir_path,
|
||||
bool make_source_readonly,
|
||||
@ -108,7 +91,7 @@ public:
|
||||
bool copy_instead_of_hardlink,
|
||||
const NameSet & files_to_copy_instead_of_hardlinks) const override;
|
||||
|
||||
DataPartStoragePtr clone(
|
||||
MutableDataPartStoragePtr clonePart(
|
||||
const std::string & to,
|
||||
const std::string & dir_path,
|
||||
const DiskPtr & disk,
|
||||
@ -116,11 +99,51 @@ public:
|
||||
|
||||
void changeRootPath(const std::string & from_root, const std::string & to_root) override;
|
||||
|
||||
DataPartStorageBuilderPtr getBuilder() const override;
|
||||
void createDirectories() override;
|
||||
void createProjection(const std::string & name) override;
|
||||
|
||||
std::unique_ptr<WriteBufferFromFileBase> writeFile(
|
||||
const String & name,
|
||||
size_t buf_size,
|
||||
const WriteSettings & settings) override;
|
||||
|
||||
std::unique_ptr<WriteBufferFromFileBase> writeTransactionFile(WriteMode mode) const override;
|
||||
|
||||
void createFile(const String & name) override;
|
||||
void moveFile(const String & from_name, const String & to_name) override;
|
||||
void replaceFile(const String & from_name, const String & to_name) override;
|
||||
|
||||
void removeFile(const String & name) override;
|
||||
void removeFileIfExists(const String & name) override;
|
||||
void removeRecursive() override;
|
||||
void removeSharedRecursive(bool keep_in_remote_fs) override;
|
||||
|
||||
SyncGuardPtr getDirectorySyncGuard() const override;
|
||||
|
||||
void createHardLinkFrom(const IDataPartStorage & source, const std::string & from, const std::string & to) override;
|
||||
|
||||
void rename(
|
||||
const std::string & new_root_path,
|
||||
const std::string & new_part_dir,
|
||||
Poco::Logger * log,
|
||||
bool remove_new_dir_if_exists,
|
||||
bool fsync_part_dir) override;
|
||||
|
||||
void beginTransaction() override;
|
||||
void commitTransaction() override;
|
||||
bool hasActiveTransaction() const override { return transaction != nullptr; }
|
||||
|
||||
private:
|
||||
VolumePtr volume;
|
||||
std::string root_path;
|
||||
std::string part_dir;
|
||||
DiskTransactionPtr transaction;
|
||||
bool has_shared_transaction = false;
|
||||
|
||||
DataPartStorageOnDisk(VolumePtr volume_, std::string root_path_, std::string part_dir_, DiskTransactionPtr transaction_);
|
||||
|
||||
template <typename Op>
|
||||
void executeOperation(Op && op);
|
||||
|
||||
void clearDirectory(
|
||||
const std::string & dir,
|
||||
@ -134,56 +157,4 @@ private:
|
||||
bool is_projection) const;
|
||||
};
|
||||
|
||||
class DataPartStorageBuilderOnDisk final : public IDataPartStorageBuilder
|
||||
{
|
||||
public:
|
||||
DataPartStorageBuilderOnDisk(VolumePtr volume_, std::string root_path_, std::string part_dir_);
|
||||
|
||||
void setRelativePath(const std::string & path) override;
|
||||
|
||||
bool exists() const override;
|
||||
|
||||
void createDirectories() override;
|
||||
void createProjection(const std::string & name) override;
|
||||
|
||||
std::string getPartDirectory() const override { return part_dir; }
|
||||
std::string getFullPath() const override;
|
||||
std::string getRelativePath() const override;
|
||||
|
||||
std::unique_ptr<WriteBufferFromFileBase> writeFile(
|
||||
const String & name,
|
||||
size_t buf_size,
|
||||
const WriteSettings & settings) override;
|
||||
|
||||
void removeFile(const String & name) override;
|
||||
void removeFileIfExists(const String & name) override;
|
||||
void removeRecursive() override;
|
||||
void removeSharedRecursive(bool keep_in_remote_fs) override;
|
||||
|
||||
SyncGuardPtr getDirectorySyncGuard() const override;
|
||||
|
||||
void createHardLinkFrom(const IDataPartStorage & source, const std::string & from, const std::string & to) const override;
|
||||
|
||||
ReservationPtr reserve(UInt64 bytes) override;
|
||||
|
||||
DataPartStorageBuilderPtr getProjection(const std::string & name) const override;
|
||||
|
||||
DataPartStoragePtr getStorage() const override;
|
||||
|
||||
void rename(
|
||||
const std::string & new_root_path,
|
||||
const std::string & new_part_dir,
|
||||
Poco::Logger * log,
|
||||
bool remove_new_dir_if_exists,
|
||||
bool fsync_part_dir) override;
|
||||
|
||||
void commit() override;
|
||||
|
||||
private:
|
||||
VolumePtr volume;
|
||||
std::string root_path;
|
||||
std::string part_dir;
|
||||
DiskTransactionPtr transaction;
|
||||
};
|
||||
|
||||
}
|
||||
|
||||
@ -13,9 +13,9 @@
|
||||
#include <Storages/MergeTree/MergedBlockOutputStream.h>
|
||||
#include <Storages/MergeTree/ReplicatedFetchList.h>
|
||||
#include <Storages/StorageReplicatedMergeTree.h>
|
||||
#include <Storages/MergeTree/DataPartStorageOnDisk.h>
|
||||
#include <Common/CurrentMetrics.h>
|
||||
#include <Common/NetException.h>
|
||||
#include <Storages/MergeTree/DataPartStorageOnDisk.h>
|
||||
#include <Disks/IO/createReadBufferFromFileBase.h>
|
||||
#include <base/scope_guard.h>
|
||||
#include <Poco/Net/HTTPRequest.h>
|
||||
@ -147,12 +147,13 @@ void Service::processQuery(const HTMLForm & params, ReadBuffer & /*body*/, Write
|
||||
|
||||
CurrentMetrics::Increment metric_increment{CurrentMetrics::ReplicatedSend};
|
||||
|
||||
if (part->data_part_storage->isStoredOnRemoteDisk())
|
||||
if (part->getDataPartStorage().isStoredOnRemoteDisk())
|
||||
{
|
||||
UInt64 revision = parse<UInt64>(params.get("disk_revision", "0"));
|
||||
if (revision)
|
||||
part->data_part_storage->syncRevision(revision);
|
||||
revision = part->data_part_storage->getRevision();
|
||||
part->getDataPartStorage().syncRevision(revision);
|
||||
|
||||
revision = part->getDataPartStorage().getRevision();
|
||||
if (revision)
|
||||
response.addCookie({"disk_revision", toString(revision)});
|
||||
}
|
||||
@ -184,8 +185,8 @@ void Service::processQuery(const HTMLForm & params, ReadBuffer & /*body*/, Write
|
||||
!isInMemoryPart(part) &&
|
||||
client_protocol_version >= REPLICATION_PROTOCOL_VERSION_WITH_PARTS_ZERO_COPY)
|
||||
{
|
||||
auto disk_type = part->data_part_storage->getDiskType();
|
||||
if (part->data_part_storage->supportZeroCopyReplication() && std::find(capability.begin(), capability.end(), disk_type) != capability.end())
|
||||
auto disk_type = part->getDataPartStorage().getDiskType();
|
||||
if (part->getDataPartStorage().supportZeroCopyReplication() && std::find(capability.begin(), capability.end(), disk_type) != capability.end())
|
||||
{
|
||||
/// Send metadata if the receiver's capability covers the source disk type.
|
||||
response.addCookie({"remote_fs_metadata", disk_type});
|
||||
@ -307,12 +308,12 @@ MergeTreeData::DataPart::Checksums Service::sendPartFromDisk(
|
||||
{
|
||||
String file_name = it.first;
|
||||
|
||||
UInt64 size = part->data_part_storage->getFileSize(file_name);
|
||||
UInt64 size = part->getDataPartStorage().getFileSize(file_name);
|
||||
|
||||
writeStringBinary(it.first, out);
|
||||
writeBinary(size, out);
|
||||
|
||||
auto file_in = part->data_part_storage->readFile(file_name, {}, std::nullopt, std::nullopt);
|
||||
auto file_in = part->getDataPartStorage().readFile(file_name, {}, std::nullopt, std::nullopt);
|
||||
HashingWriteBuffer hashing_out(out);
|
||||
copyDataWithThrottler(*file_in, hashing_out, blocker.getCounter(), data.getSendsThrottler());
|
||||
|
||||
@ -323,7 +324,7 @@ MergeTreeData::DataPart::Checksums Service::sendPartFromDisk(
|
||||
throw Exception(
|
||||
ErrorCodes::BAD_SIZE_OF_FILE_IN_DATA_PART,
|
||||
"Unexpected size of file {}, expected {} got {}",
|
||||
std::string(fs::path(part->data_part_storage->getRelativePath()) / file_name),
|
||||
std::string(fs::path(part->getDataPartStorage().getRelativePath()) / file_name),
|
||||
hashing_out.count(), size);
|
||||
|
||||
writePODBinary(hashing_out.getHash(), out);
|
||||
@ -342,9 +343,9 @@ MergeTreeData::DataPart::Checksums Service::sendPartFromDiskRemoteMeta(
|
||||
bool send_part_id,
|
||||
const std::map<String, std::shared_ptr<IMergeTreeDataPart>> & projections)
|
||||
{
|
||||
const auto * data_part_storage_on_disk = dynamic_cast<const DataPartStorageOnDisk *>(part->data_part_storage.get());
|
||||
const auto * data_part_storage_on_disk = dynamic_cast<const DataPartStorageOnDisk *>(&part->getDataPartStorage());
|
||||
if (!data_part_storage_on_disk)
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR, "Storage '{}' doesn't support zero-copy replication", part->data_part_storage->getDiskName());
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR, "Storage '{}' doesn't support zero-copy replication", part->getDataPartStorage().getDiskName());
|
||||
|
||||
if (!data_part_storage_on_disk->supportZeroCopyReplication())
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR, "Disk '{}' doesn't support zero-copy replication", data_part_storage_on_disk->getDiskName());
|
||||
@ -365,7 +366,7 @@ MergeTreeData::DataPart::Checksums Service::sendPartFromDiskRemoteMeta(
|
||||
std::vector<std::string> paths;
|
||||
paths.reserve(checksums.files.size());
|
||||
for (const auto & it : checksums.files)
|
||||
paths.push_back(fs::path(part->data_part_storage->getRelativePath()) / it.first);
|
||||
paths.push_back(fs::path(part->getDataPartStorage().getRelativePath()) / it.first);
|
||||
|
||||
/// Serialized metadatadatas with zero ref counts.
|
||||
auto metadatas = data_part_storage_on_disk->getSerializedMetadata(paths);
|
||||
@ -399,7 +400,7 @@ MergeTreeData::DataPart::Checksums Service::sendPartFromDiskRemoteMeta(
|
||||
for (const auto & it : checksums.files)
|
||||
{
|
||||
const String & file_name = it.first;
|
||||
String file_path_prefix = fs::path(part->data_part_storage->getRelativePath()) / file_name;
|
||||
String file_path_prefix = fs::path(part->getDataPartStorage().getRelativePath()) / file_name;
|
||||
|
||||
/// Just some additional checks
|
||||
String metadata_file_path = fs::path(data_part_storage_on_disk->getDiskPath()) / file_path_prefix;
|
||||
@ -728,13 +729,9 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToMemory(
|
||||
data.getRelativeDataPath(),
|
||||
part_name);
|
||||
|
||||
auto data_part_storage_builder = std::make_shared<DataPartStorageBuilderOnDisk>(
|
||||
volume,
|
||||
data.getRelativeDataPath(),
|
||||
part_name);
|
||||
|
||||
MergeTreeData::MutableDataPartPtr new_data_part =
|
||||
std::make_shared<MergeTreeDataPartInMemory>(data, part_name, data_part_storage);
|
||||
|
||||
new_data_part->version.setCreationTID(Tx::PrehistoricTID, nullptr);
|
||||
|
||||
for (auto i = 0ul; i < projections; ++i)
|
||||
@ -750,7 +747,6 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToMemory(
|
||||
throttler->add(block.bytes());
|
||||
|
||||
auto projection_part_storage = data_part_storage->getProjection(projection_name + ".proj");
|
||||
auto projection_part_storage_builder = data_part_storage_builder->getProjection(projection_name + ".proj");
|
||||
|
||||
MergeTreePartInfo new_part_info("all", 0, 0, 0);
|
||||
MergeTreeData::MutableDataPartPtr new_projection_part =
|
||||
@ -764,7 +760,6 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToMemory(
|
||||
|
||||
MergedBlockOutputStream part_out(
|
||||
new_projection_part,
|
||||
projection_part_storage_builder,
|
||||
metadata_snapshot->projections.get(projection_name).metadata,
|
||||
block.getNamesAndTypesList(),
|
||||
{},
|
||||
@ -792,7 +787,7 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToMemory(
|
||||
new_data_part->partition.create(metadata_snapshot, block, 0, context);
|
||||
|
||||
MergedBlockOutputStream part_out(
|
||||
new_data_part, data_part_storage_builder, metadata_snapshot, block.getNamesAndTypesList(), {},
|
||||
new_data_part, metadata_snapshot, block.getNamesAndTypesList(), {},
|
||||
CompressionCodecFactory::instance().get("NONE", {}), NO_TRANSACTION_PTR);
|
||||
|
||||
part_out.write(block);
|
||||
@ -804,7 +799,7 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToMemory(
|
||||
|
||||
void Fetcher::downloadBasePartOrProjectionPartToDiskRemoteMeta(
|
||||
const String & replica_path,
|
||||
DataPartStorageBuilderPtr & data_part_storage_builder,
|
||||
const MutableDataPartStoragePtr & data_part_storage,
|
||||
PooledReadWriteBufferFromHTTP & in,
|
||||
MergeTreeData::DataPart::Checksums & checksums,
|
||||
ThrottlerPtr throttler) const
|
||||
@ -820,7 +815,7 @@ void Fetcher::downloadBasePartOrProjectionPartToDiskRemoteMeta(
|
||||
readStringBinary(file_name, in);
|
||||
readBinary(file_size, in);
|
||||
|
||||
String metadata_file = fs::path(data_part_storage_builder->getFullPath()) / file_name;
|
||||
String metadata_file = fs::path(data_part_storage->getFullPath()) / file_name;
|
||||
|
||||
{
|
||||
auto file_out = std::make_unique<WriteBufferFromFile>(metadata_file, DBMS_DEFAULT_BUFFER_SIZE, -1, 0666, nullptr, 0);
|
||||
@ -834,8 +829,8 @@ void Fetcher::downloadBasePartOrProjectionPartToDiskRemoteMeta(
|
||||
/// NOTE The is_cancelled flag also makes sense to check every time you read over the network,
|
||||
/// performing a poll with a not very large timeout.
|
||||
/// And now we check it only between read chunks (in the `copyData` function).
|
||||
data_part_storage_builder->removeSharedRecursive(true);
|
||||
data_part_storage_builder->commit();
|
||||
data_part_storage->removeSharedRecursive(true);
|
||||
data_part_storage->commitTransaction();
|
||||
throw Exception("Fetching of part was cancelled", ErrorCodes::ABORTED);
|
||||
}
|
||||
|
||||
@ -861,7 +856,7 @@ void Fetcher::downloadBasePartOrProjectionPartToDiskRemoteMeta(
|
||||
|
||||
void Fetcher::downloadBaseOrProjectionPartToDisk(
|
||||
const String & replica_path,
|
||||
DataPartStorageBuilderPtr & data_part_storage_builder,
|
||||
const MutableDataPartStoragePtr & data_part_storage,
|
||||
bool sync,
|
||||
PooledReadWriteBufferFromHTTP & in,
|
||||
MergeTreeData::DataPart::Checksums & checksums,
|
||||
@ -880,14 +875,14 @@ void Fetcher::downloadBaseOrProjectionPartToDisk(
|
||||
|
||||
/// File must be inside "absolute_part_path" directory.
|
||||
/// Otherwise malicious ClickHouse replica may force us to write to arbitrary path.
|
||||
String absolute_file_path = fs::weakly_canonical(fs::path(data_part_storage_builder->getRelativePath()) / file_name);
|
||||
if (!startsWith(absolute_file_path, fs::weakly_canonical(data_part_storage_builder->getRelativePath()).string()))
|
||||
String absolute_file_path = fs::weakly_canonical(fs::path(data_part_storage->getRelativePath()) / file_name);
|
||||
if (!startsWith(absolute_file_path, fs::weakly_canonical(data_part_storage->getRelativePath()).string()))
|
||||
throw Exception(ErrorCodes::INSECURE_PATH,
|
||||
"File path ({}) doesn't appear to be inside part path ({}). "
|
||||
"This may happen if we are trying to download part from malicious replica or logical error.",
|
||||
absolute_file_path, data_part_storage_builder->getRelativePath());
|
||||
absolute_file_path, data_part_storage->getRelativePath());
|
||||
|
||||
auto file_out = data_part_storage_builder->writeFile(file_name, std::min<UInt64>(file_size, DBMS_DEFAULT_BUFFER_SIZE), {});
|
||||
auto file_out = data_part_storage->writeFile(file_name, std::min<UInt64>(file_size, DBMS_DEFAULT_BUFFER_SIZE), {});
|
||||
HashingWriteBuffer hashing_out(*file_out);
|
||||
copyDataWithThrottler(in, hashing_out, file_size, blocker.getCounter(), throttler);
|
||||
|
||||
@ -896,7 +891,7 @@ void Fetcher::downloadBaseOrProjectionPartToDisk(
|
||||
/// NOTE The is_cancelled flag also makes sense to check every time you read over the network,
|
||||
/// performing a poll with a not very large timeout.
|
||||
/// And now we check it only between read chunks (in the `copyData` function).
|
||||
data_part_storage_builder->removeRecursive();
|
||||
data_part_storage->removeRecursive();
|
||||
throw Exception("Fetching of part was cancelled", ErrorCodes::ABORTED);
|
||||
}
|
||||
|
||||
@ -906,7 +901,7 @@ void Fetcher::downloadBaseOrProjectionPartToDisk(
|
||||
if (expected_hash != hashing_out.getHash())
|
||||
throw Exception(ErrorCodes::CHECKSUM_DOESNT_MATCH,
|
||||
"Checksum mismatch for file {} transferred from {}",
|
||||
(fs::path(data_part_storage_builder->getFullPath()) / file_name).string(),
|
||||
(fs::path(data_part_storage->getFullPath()) / file_name).string(),
|
||||
replica_path);
|
||||
|
||||
if (file_name != "checksums.txt" &&
|
||||
@ -951,15 +946,12 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToDisk(
|
||||
part_relative_path,
|
||||
part_dir);
|
||||
|
||||
DataPartStorageBuilderPtr data_part_storage_builder = std::make_shared<DataPartStorageBuilderOnDisk>(
|
||||
volume,
|
||||
part_relative_path,
|
||||
part_dir);
|
||||
data_part_storage->beginTransaction();
|
||||
|
||||
if (data_part_storage_builder->exists())
|
||||
if (data_part_storage->exists())
|
||||
{
|
||||
LOG_WARNING(log, "Directory {} already exists, probably result of a failed fetch. Will remove it before fetching part.",
|
||||
data_part_storage_builder->getFullPath());
|
||||
data_part_storage->getFullPath());
|
||||
|
||||
/// Even if it's a temporary part it could be downloaded with zero copy replication and this function
|
||||
/// is executed as a callback.
|
||||
@ -967,10 +959,10 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToDisk(
|
||||
/// We don't control the amount of refs for temporary parts so we cannot decide can we remove blobs
|
||||
/// or not. So we are not doing it
|
||||
bool keep_shared = disk->supportZeroCopyReplication() && data_settings->allow_remote_fs_zero_copy_replication;
|
||||
data_part_storage_builder->removeSharedRecursive(keep_shared);
|
||||
data_part_storage->removeSharedRecursive(keep_shared);
|
||||
}
|
||||
|
||||
data_part_storage_builder->createDirectories();
|
||||
data_part_storage->createDirectories();
|
||||
|
||||
SyncGuardPtr sync_guard;
|
||||
if (data.getSettings()->fsync_part_directory)
|
||||
@ -985,19 +977,18 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToDisk(
|
||||
MergeTreeData::DataPart::Checksums projection_checksum;
|
||||
|
||||
auto projection_part_storage = data_part_storage->getProjection(projection_name + ".proj");
|
||||
auto projection_part_storage_builder = data_part_storage_builder->getProjection(projection_name + ".proj");
|
||||
|
||||
projection_part_storage_builder->createDirectories();
|
||||
projection_part_storage->createDirectories();
|
||||
downloadBaseOrProjectionPartToDisk(
|
||||
replica_path, projection_part_storage_builder, sync, in, projection_checksum, throttler);
|
||||
replica_path, projection_part_storage, sync, in, projection_checksum, throttler);
|
||||
checksums.addFile(
|
||||
projection_name + ".proj", projection_checksum.getTotalSizeOnDisk(), projection_checksum.getTotalChecksumUInt128());
|
||||
}
|
||||
|
||||
// Download the base part
|
||||
downloadBaseOrProjectionPartToDisk(replica_path, data_part_storage_builder, sync, in, checksums, throttler);
|
||||
downloadBaseOrProjectionPartToDisk(replica_path, data_part_storage, sync, in, checksums, throttler);
|
||||
|
||||
assertEOF(in);
|
||||
data_part_storage->commitTransaction();
|
||||
MergeTreeData::MutableDataPartPtr new_data_part = data.createPart(part_name, data_part_storage);
|
||||
new_data_part->version.setCreationTID(Tx::PrehistoricTID, nullptr);
|
||||
new_data_part->is_temp = true;
|
||||
@ -1043,17 +1034,14 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToDiskRemoteMeta(
|
||||
part_relative_path,
|
||||
part_dir);
|
||||
|
||||
DataPartStorageBuilderPtr data_part_storage_builder = std::make_shared<DataPartStorageBuilderOnDisk>(
|
||||
volume,
|
||||
part_relative_path,
|
||||
part_dir);
|
||||
data_part_storage->beginTransaction();
|
||||
|
||||
if (data_part_storage->exists())
|
||||
throw Exception(ErrorCodes::DIRECTORY_ALREADY_EXISTS, "Directory {} already exists.", data_part_storage->getFullPath());
|
||||
|
||||
CurrentMetrics::Increment metric_increment{CurrentMetrics::ReplicatedFetch};
|
||||
|
||||
volume->getDisk()->createDirectories(data_part_storage->getFullPath());
|
||||
data_part_storage->createDirectories();
|
||||
|
||||
for (auto i = 0ul; i < projections; ++i)
|
||||
{
|
||||
@ -1062,24 +1050,22 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToDiskRemoteMeta(
|
||||
MergeTreeData::DataPart::Checksums projection_checksum;
|
||||
|
||||
auto projection_part_storage = data_part_storage->getProjection(projection_name + ".proj");
|
||||
auto projection_part_storage_builder = data_part_storage_builder->getProjection(projection_name + ".proj");
|
||||
|
||||
projection_part_storage_builder->createDirectories();
|
||||
projection_part_storage->createDirectories();
|
||||
downloadBasePartOrProjectionPartToDiskRemoteMeta(
|
||||
replica_path, projection_part_storage_builder, in, projection_checksum, throttler);
|
||||
replica_path, projection_part_storage, in, projection_checksum, throttler);
|
||||
|
||||
checksums.addFile(
|
||||
projection_name + ".proj", projection_checksum.getTotalSizeOnDisk(), projection_checksum.getTotalChecksumUInt128());
|
||||
}
|
||||
|
||||
downloadBasePartOrProjectionPartToDiskRemoteMeta(
|
||||
replica_path, data_part_storage_builder, in, checksums, throttler);
|
||||
replica_path, data_part_storage, in, checksums, throttler);
|
||||
|
||||
assertEOF(in);
|
||||
MergeTreeData::MutableDataPartPtr new_data_part;
|
||||
try
|
||||
{
|
||||
data_part_storage_builder->commit();
|
||||
data_part_storage->commitTransaction();
|
||||
|
||||
new_data_part = data.createPart(part_name, data_part_storage);
|
||||
new_data_part->version.setCreationTID(Tx::PrehistoricTID, nullptr);
|
||||
|
||||
@ -94,7 +94,7 @@ public:
|
||||
private:
|
||||
void downloadBaseOrProjectionPartToDisk(
|
||||
const String & replica_path,
|
||||
DataPartStorageBuilderPtr & data_part_storage_builder,
|
||||
const MutableDataPartStoragePtr & data_part_storage,
|
||||
bool sync,
|
||||
PooledReadWriteBufferFromHTTP & in,
|
||||
MergeTreeData::DataPart::Checksums & checksums,
|
||||
@ -102,12 +102,11 @@ private:
|
||||
|
||||
void downloadBasePartOrProjectionPartToDiskRemoteMeta(
|
||||
const String & replica_path,
|
||||
DataPartStorageBuilderPtr & data_part_storage_builder,
|
||||
const MutableDataPartStoragePtr & data_part_storage,
|
||||
PooledReadWriteBufferFromHTTP & in,
|
||||
MergeTreeData::DataPart::Checksums & checksums,
|
||||
ThrottlerPtr throttler) const;
|
||||
|
||||
|
||||
MergeTreeData::MutableDataPartPtr downloadPartToDisk(
|
||||
const String & part_name,
|
||||
const String & replica_path,
|
||||
|
||||
@ -4,6 +4,9 @@
|
||||
#include <Core/NamesAndTypes.h>
|
||||
#include <Interpreters/TransactionVersionMetadata.h>
|
||||
#include <Storages/MergeTree/MergeTreeDataPartState.h>
|
||||
#include <Disks/WriteMode.h>
|
||||
#include <boost/core/noncopyable.hpp>
|
||||
#include <memory>
|
||||
#include <optional>
|
||||
|
||||
namespace DB
|
||||
@ -18,6 +21,7 @@ struct CanRemoveDescription
|
||||
NameSet files_not_to_remove;
|
||||
|
||||
};
|
||||
|
||||
using CanRemoveCallback = std::function<CanRemoveDescription()>;
|
||||
|
||||
class IDataPartStorageIterator
|
||||
@ -61,13 +65,10 @@ struct WriteSettings;
|
||||
|
||||
class TemporaryFileOnDisk;
|
||||
|
||||
class IDataPartStorageBuilder;
|
||||
using DataPartStorageBuilderPtr = std::shared_ptr<IDataPartStorageBuilder>;
|
||||
|
||||
/// This is an abstraction of storage for data part files.
|
||||
/// Ideally, it is assumed to contains read-only methods from IDisk.
|
||||
/// It is not fulfilled now, but let's try our best.
|
||||
class IDataPartStorage
|
||||
class IDataPartStorage : public boost::noncopyable
|
||||
{
|
||||
public:
|
||||
virtual ~IDataPartStorage() = default;
|
||||
@ -81,16 +82,19 @@ public:
|
||||
/// virtual std::string getRelativeRootPath() const = 0;
|
||||
|
||||
/// Get a storage for projection.
|
||||
virtual std::shared_ptr<IDataPartStorage> getProjection(const std::string & name) const = 0;
|
||||
virtual std::shared_ptr<IDataPartStorage> getProjection(const std::string & name) = 0;
|
||||
virtual std::shared_ptr<const IDataPartStorage> getProjection(const std::string & name) const = 0;
|
||||
|
||||
/// Part directory exists.
|
||||
virtual bool exists() const = 0;
|
||||
|
||||
/// File inside part directory exists. Specified path is relative to the part path.
|
||||
virtual bool exists(const std::string & name) const = 0;
|
||||
virtual bool isDirectory(const std::string & name) const = 0;
|
||||
|
||||
/// Modification time for part directory.
|
||||
virtual Poco::Timestamp getLastModified() const = 0;
|
||||
|
||||
/// Iterate part directory. Iteration in subdirectory is not needed yet.
|
||||
virtual DataPartStorageIteratorPtr iterate() const = 0;
|
||||
|
||||
@ -107,7 +111,6 @@ public:
|
||||
std::optional<size_t> read_hint,
|
||||
std::optional<size_t> file_size) const = 0;
|
||||
|
||||
virtual void loadVersionMetadata(VersionMetadata & version, Poco::Logger * log) const = 0;
|
||||
virtual void checkConsistency(const MergeTreeDataPartChecksums & checksums) const = 0;
|
||||
|
||||
struct ProjectionChecksums
|
||||
@ -129,12 +132,12 @@ public:
|
||||
|
||||
/// Get a name like 'prefix_partdir_tryN' which does not exist in a root dir.
|
||||
/// TODO: remove it.
|
||||
virtual std::optional<String> getRelativePathForPrefix(Poco::Logger * log, const String & prefix, bool detached, bool broken) const = 0;
|
||||
virtual std::optional<String> getRelativePathForPrefix(
|
||||
Poco::Logger * log, const String & prefix, bool detached, bool broken) const = 0;
|
||||
|
||||
/// Reset part directory, used for im-memory parts.
|
||||
/// Reset part directory, used for in-memory parts.
|
||||
/// TODO: remove it.
|
||||
virtual void setRelativePath(const std::string & path) = 0;
|
||||
virtual void onRename(const std::string & new_root_path, const std::string & new_part_dir) = 0;
|
||||
|
||||
/// Some methods from IDisk. Needed to avoid getting internal IDisk interface.
|
||||
virtual std::string getDiskName() const = 0;
|
||||
@ -143,41 +146,26 @@ public:
|
||||
virtual bool supportZeroCopyReplication() const { return false; }
|
||||
virtual bool supportParallelWrite() const = 0;
|
||||
virtual bool isBroken() const = 0;
|
||||
virtual void syncRevision(UInt64 revision) = 0;
|
||||
|
||||
/// TODO: remove or at least remove const.
|
||||
virtual void syncRevision(UInt64 revision) const = 0;
|
||||
virtual UInt64 getRevision() const = 0;
|
||||
|
||||
virtual std::unordered_map<String, String> getSerializedMetadata(const std::vector<String> & paths) const = 0;
|
||||
/// Get a path for internal disk if relevant. It is used mainly for logging.
|
||||
virtual std::string getDiskPath() const = 0;
|
||||
|
||||
/// Check if data part is stored on one of the specified disk in set.
|
||||
using DisksSet = std::unordered_set<DiskPtr>;
|
||||
virtual DisksSet::const_iterator isStoredOnDisk(const DisksSet & disks) const { return disks.end(); }
|
||||
|
||||
/// Reserve space on the same disk.
|
||||
/// Probably we should try to remove it later.
|
||||
virtual ReservationPtr reserve(UInt64 /*bytes*/) const { return nullptr; }
|
||||
virtual ReservationPtr tryReserve(UInt64 /*bytes*/) const { return nullptr; }
|
||||
virtual size_t getVolumeIndex(const IStoragePolicy &) const { return 0; }
|
||||
|
||||
/// Some methods which change data part internals possibly after creation.
|
||||
/// Probably we should try to remove it later.
|
||||
virtual void writeChecksums(const MergeTreeDataPartChecksums & checksums, const WriteSettings & settings) const = 0;
|
||||
virtual void writeColumns(const NamesAndTypesList & columns, const WriteSettings & settings) const = 0;
|
||||
virtual void writeVersionMetadata(const VersionMetadata & version, bool fsync_part_dir) const = 0;
|
||||
virtual void appendCSNToVersionMetadata(const VersionMetadata & version, VersionMetadata::WhichCSN which_csn) const = 0;
|
||||
virtual void appendRemovalTIDToVersionMetadata(const VersionMetadata & version, bool clear) const = 0;
|
||||
virtual void writeDeleteOnDestroyMarker(Poco::Logger * log) const = 0;
|
||||
virtual void removeDeleteOnDestroyMarker() const = 0;
|
||||
virtual void removeVersionMetadata() const = 0;
|
||||
/// TODO: remove constness
|
||||
virtual ReservationPtr reserve(UInt64 /*bytes*/) const { return nullptr; }
|
||||
virtual ReservationPtr tryReserve(UInt64 /*bytes*/) const { return nullptr; }
|
||||
|
||||
/// A leak of abstraction.
|
||||
/// Return some uniq string for file.
|
||||
/// Required for distinguish different copies of the same part on remote FS.
|
||||
virtual String getUniqueId() const = 0;
|
||||
|
||||
/// A leak of abstraction
|
||||
virtual bool shallParticipateInMerges(const IStoragePolicy &) const { return true; }
|
||||
|
||||
/// Create a backup of a data part.
|
||||
/// This method adds a new entry to backup_entries.
|
||||
/// Also creates a new tmp_dir for internal disk (if disk is mentioned the first time).
|
||||
@ -205,7 +193,7 @@ public:
|
||||
const NameSet & files_to_copy_instead_of_hardlinks) const = 0;
|
||||
|
||||
/// Make a full copy of a data part into 'to/dir_path' (possibly to a different disk).
|
||||
virtual std::shared_ptr<IDataPartStorage> clone(
|
||||
virtual std::shared_ptr<IDataPartStorage> clonePart(
|
||||
const std::string & to,
|
||||
const std::string & dir_path,
|
||||
const DiskPtr & disk,
|
||||
@ -215,33 +203,22 @@ public:
|
||||
/// Right now, this is needed for rename table query.
|
||||
virtual void changeRootPath(const std::string & from_root, const std::string & to_root) = 0;
|
||||
|
||||
/// Leak of abstraction as well. We should use builder as one-time object which allow
|
||||
/// us to build parts, while storage should be read-only method to access part properties
|
||||
/// related to disk. However our code is really tricky and sometimes we need ad-hoc builders.
|
||||
virtual DataPartStorageBuilderPtr getBuilder() const = 0;
|
||||
};
|
||||
|
||||
using DataPartStoragePtr = std::shared_ptr<IDataPartStorage>;
|
||||
|
||||
/// This interface is needed to write data part.
|
||||
class IDataPartStorageBuilder
|
||||
{
|
||||
public:
|
||||
virtual ~IDataPartStorageBuilder() = default;
|
||||
|
||||
/// Reset part directory, used for im-memory parts
|
||||
virtual void setRelativePath(const std::string & path) = 0;
|
||||
|
||||
virtual std::string getPartDirectory() const = 0;
|
||||
virtual std::string getFullPath() const = 0;
|
||||
virtual std::string getRelativePath() const = 0;
|
||||
|
||||
virtual bool exists() const = 0;
|
||||
|
||||
virtual void createDirectories() = 0;
|
||||
virtual void createProjection(const std::string & name) = 0;
|
||||
|
||||
virtual std::unique_ptr<WriteBufferFromFileBase> writeFile(const String & name, size_t buf_size, const WriteSettings & settings) = 0;
|
||||
virtual std::unique_ptr<WriteBufferFromFileBase> writeFile(
|
||||
const String & name,
|
||||
size_t buf_size,
|
||||
const WriteSettings & settings) = 0;
|
||||
|
||||
/// A special const method to write transaction file.
|
||||
/// It's const, because file with transaction metadata
|
||||
/// can be modified after part creation.
|
||||
virtual std::unique_ptr<WriteBufferFromFileBase> writeTransactionFile(WriteMode mode) const = 0;
|
||||
|
||||
virtual void createFile(const String & name) = 0;
|
||||
virtual void moveFile(const String & from_name, const String & to_name) = 0;
|
||||
virtual void replaceFile(const String & from_name, const String & to_name) = 0;
|
||||
|
||||
virtual void removeFile(const String & name) = 0;
|
||||
virtual void removeFileIfExists(const String & name) = 0;
|
||||
@ -250,20 +227,12 @@ public:
|
||||
|
||||
virtual SyncGuardPtr getDirectorySyncGuard() const { return nullptr; }
|
||||
|
||||
virtual void createHardLinkFrom(const IDataPartStorage & source, const std::string & from, const std::string & to) const = 0;
|
||||
|
||||
virtual ReservationPtr reserve(UInt64 /*bytes*/) { return nullptr; }
|
||||
|
||||
virtual std::shared_ptr<IDataPartStorageBuilder> getProjection(const std::string & name) const = 0;
|
||||
|
||||
virtual DataPartStoragePtr getStorage() const = 0;
|
||||
virtual void createHardLinkFrom(const IDataPartStorage & source, const std::string & from, const std::string & to) = 0;
|
||||
|
||||
/// Rename part.
|
||||
/// Ideally, new_root_path should be the same as current root (but it is not true).
|
||||
/// Examples are: 'all_1_2_1' -> 'detached/all_1_2_1'
|
||||
/// 'moving/tmp_all_1_2_1' -> 'all_1_2_1'
|
||||
///
|
||||
/// To notify storage also call onRename for it with first two args
|
||||
virtual void rename(
|
||||
const std::string & new_root_path,
|
||||
const std::string & new_part_dir,
|
||||
@ -271,7 +240,35 @@ public:
|
||||
bool remove_new_dir_if_exists,
|
||||
bool fsync_part_dir) = 0;
|
||||
|
||||
virtual void commit() = 0;
|
||||
/// Starts a transaction of mutable operations.
|
||||
virtual void beginTransaction() = 0;
|
||||
/// Commits a transaction of mutable operations.
|
||||
virtual void commitTransaction() = 0;
|
||||
virtual bool hasActiveTransaction() const = 0;
|
||||
};
|
||||
|
||||
using DataPartStoragePtr = std::shared_ptr<const IDataPartStorage>;
|
||||
using MutableDataPartStoragePtr = std::shared_ptr<IDataPartStorage>;
|
||||
|
||||
/// A holder that encapsulates data part storage and
|
||||
/// gives access to const storage from const methods
|
||||
/// and to mutable storage from non-const methods.
|
||||
class DataPartStorageHolder : public boost::noncopyable
|
||||
{
|
||||
public:
|
||||
explicit DataPartStorageHolder(MutableDataPartStoragePtr storage_)
|
||||
: storage(std::move(storage_))
|
||||
{
|
||||
}
|
||||
|
||||
IDataPartStorage & getDataPartStorage() { return *storage; }
|
||||
const IDataPartStorage & getDataPartStorage() const { return *storage; }
|
||||
|
||||
MutableDataPartStoragePtr getDataPartStoragePtr() { return storage; }
|
||||
DataPartStoragePtr getDataPartStoragePtr() const { return storage; }
|
||||
|
||||
private:
|
||||
MutableDataPartStoragePtr storage;
|
||||
};
|
||||
|
||||
}
|
||||
|
||||
@ -1,4 +1,5 @@
|
||||
#include "IMergeTreeDataPart.h"
|
||||
#include "Storages/MergeTree/IDataPartStorage.h"
|
||||
|
||||
#include <optional>
|
||||
#include <boost/algorithm/string/join.hpp>
|
||||
@ -101,7 +102,7 @@ void IMergeTreeDataPart::MinMaxIndex::load(const MergeTreeData & data, const Par
|
||||
}
|
||||
|
||||
IMergeTreeDataPart::MinMaxIndex::WrittenFiles IMergeTreeDataPart::MinMaxIndex::store(
|
||||
const MergeTreeData & data, const DataPartStorageBuilderPtr & data_part_storage_builder, Checksums & out_checksums) const
|
||||
const MergeTreeData & data, IDataPartStorage & part_storage, Checksums & out_checksums) const
|
||||
{
|
||||
auto metadata_snapshot = data.getInMemoryMetadataPtr();
|
||||
const auto & partition_key = metadata_snapshot->getPartitionKey();
|
||||
@ -109,20 +110,20 @@ IMergeTreeDataPart::MinMaxIndex::WrittenFiles IMergeTreeDataPart::MinMaxIndex::s
|
||||
auto minmax_column_names = data.getMinMaxColumnsNames(partition_key);
|
||||
auto minmax_column_types = data.getMinMaxColumnsTypes(partition_key);
|
||||
|
||||
return store(minmax_column_names, minmax_column_types, data_part_storage_builder, out_checksums);
|
||||
return store(minmax_column_names, minmax_column_types, part_storage, out_checksums);
|
||||
}
|
||||
|
||||
IMergeTreeDataPart::MinMaxIndex::WrittenFiles IMergeTreeDataPart::MinMaxIndex::store(
|
||||
const Names & column_names,
|
||||
const DataTypes & data_types,
|
||||
const DataPartStorageBuilderPtr & data_part_storage_builder,
|
||||
IDataPartStorage & part_storage,
|
||||
Checksums & out_checksums) const
|
||||
{
|
||||
if (!initialized)
|
||||
throw Exception(
|
||||
ErrorCodes::LOGICAL_ERROR,
|
||||
"Attempt to store uninitialized MinMax index for part {}. This is a bug",
|
||||
data_part_storage_builder->getFullPath());
|
||||
part_storage.getFullPath());
|
||||
|
||||
WrittenFiles written_files;
|
||||
|
||||
@ -131,7 +132,7 @@ IMergeTreeDataPart::MinMaxIndex::WrittenFiles IMergeTreeDataPart::MinMaxIndex::s
|
||||
String file_name = "minmax_" + escapeForFileName(column_names[i]) + ".idx";
|
||||
auto serialization = data_types.at(i)->getDefaultSerialization();
|
||||
|
||||
auto out = data_part_storage_builder->writeFile(file_name, DBMS_DEFAULT_BUFFER_SIZE, {});
|
||||
auto out = part_storage.writeFile(file_name, DBMS_DEFAULT_BUFFER_SIZE, {});
|
||||
HashingWriteBuffer out_hashing(*out);
|
||||
serialization->serializeBinary(hyperrectangle[i].left, out_hashing);
|
||||
serialization->serializeBinary(hyperrectangle[i].right, out_hashing);
|
||||
@ -301,13 +302,13 @@ static void decrementTypeMetric(MergeTreeDataPartType type)
|
||||
IMergeTreeDataPart::IMergeTreeDataPart(
|
||||
const MergeTreeData & storage_,
|
||||
const String & name_,
|
||||
const DataPartStoragePtr & data_part_storage_,
|
||||
const MutableDataPartStoragePtr & data_part_storage_,
|
||||
Type part_type_,
|
||||
const IMergeTreeDataPart * parent_part_)
|
||||
: storage(storage_)
|
||||
: DataPartStorageHolder(data_part_storage_)
|
||||
, storage(storage_)
|
||||
, name(name_)
|
||||
, info(MergeTreePartInfo::fromPartName(name_, storage.format_version))
|
||||
, data_part_storage(parent_part_ ? parent_part_->data_part_storage : data_part_storage_)
|
||||
, index_granularity_info(storage_, part_type_)
|
||||
, part_type(part_type_)
|
||||
, parent_part(parent_part_)
|
||||
@ -315,6 +316,7 @@ IMergeTreeDataPart::IMergeTreeDataPart(
|
||||
{
|
||||
if (parent_part)
|
||||
state = MergeTreeDataPartState::Active;
|
||||
|
||||
incrementStateMetric(state);
|
||||
incrementTypeMetric(part_type);
|
||||
|
||||
@ -328,13 +330,13 @@ IMergeTreeDataPart::IMergeTreeDataPart(
|
||||
const MergeTreeData & storage_,
|
||||
const String & name_,
|
||||
const MergeTreePartInfo & info_,
|
||||
const DataPartStoragePtr & data_part_storage_,
|
||||
const MutableDataPartStoragePtr & data_part_storage_,
|
||||
Type part_type_,
|
||||
const IMergeTreeDataPart * parent_part_)
|
||||
: storage(storage_)
|
||||
: DataPartStorageHolder(data_part_storage_)
|
||||
, storage(storage_)
|
||||
, name(name_)
|
||||
, info(info_)
|
||||
, data_part_storage(data_part_storage_)
|
||||
, index_granularity_info(storage_, part_type_)
|
||||
, part_type(part_type_)
|
||||
, parent_part(parent_part_)
|
||||
@ -342,6 +344,7 @@ IMergeTreeDataPart::IMergeTreeDataPart(
|
||||
{
|
||||
if (parent_part)
|
||||
state = MergeTreeDataPartState::Active;
|
||||
|
||||
incrementStateMetric(state);
|
||||
incrementTypeMetric(part_type);
|
||||
|
||||
@ -505,17 +508,17 @@ void IMergeTreeDataPart::removeIfNeeded()
|
||||
std::string path;
|
||||
try
|
||||
{
|
||||
path = data_part_storage->getRelativePath();
|
||||
path = getDataPartStorage().getRelativePath();
|
||||
|
||||
if (!data_part_storage->exists()) // path
|
||||
if (!getDataPartStorage().exists()) // path
|
||||
return;
|
||||
|
||||
if (is_temp)
|
||||
{
|
||||
String file_name = fileName(data_part_storage->getPartDirectory());
|
||||
String file_name = fileName(getDataPartStorage().getPartDirectory());
|
||||
|
||||
if (file_name.empty())
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR, "relative_path {} of part {} is invalid or not set", data_part_storage->getPartDirectory(), name);
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR, "relative_path {} of part {} is invalid or not set", getDataPartStorage().getPartDirectory(), name);
|
||||
|
||||
if (!startsWith(file_name, "tmp") && !endsWith(file_name, ".tmp_proj"))
|
||||
{
|
||||
@ -620,7 +623,7 @@ String IMergeTreeDataPart::getColumnNameWithMinimumCompressedSize(bool with_subc
|
||||
}
|
||||
|
||||
if (!minimum_size_column)
|
||||
throw Exception("Could not find a column of minimum size in MergeTree, part " + data_part_storage->getFullPath(), ErrorCodes::LOGICAL_ERROR);
|
||||
throw Exception("Could not find a column of minimum size in MergeTree, part " + getDataPartStorage().getFullPath(), ErrorCodes::LOGICAL_ERROR);
|
||||
|
||||
return *minimum_size_column;
|
||||
}
|
||||
@ -698,9 +701,9 @@ void IMergeTreeDataPart::loadProjections(bool require_columns_checksums, bool ch
|
||||
for (const auto & projection : metadata_snapshot->projections)
|
||||
{
|
||||
String path = /*getRelativePath() + */ projection.name + ".proj";
|
||||
if (data_part_storage->exists(path))
|
||||
if (getDataPartStorage().exists(path))
|
||||
{
|
||||
auto projection_part_storage = data_part_storage->getProjection(projection.name + ".proj");
|
||||
auto projection_part_storage = getDataPartStorage().getProjection(projection.name + ".proj");
|
||||
auto part = storage.createPart(projection.name, {"all", 0, 0, 0}, projection_part_storage, this);
|
||||
part->loadColumnsChecksumsIndexes(require_columns_checksums, check_consistency);
|
||||
projection_parts.emplace(projection.name, std::move(part));
|
||||
@ -741,8 +744,8 @@ void IMergeTreeDataPart::loadIndex()
|
||||
loaded_index[i]->reserve(index_granularity.getMarksCount());
|
||||
}
|
||||
|
||||
String index_name = "primary" + getIndexExtensionFromFilesystem(data_part_storage).value();
|
||||
String index_path = fs::path(data_part_storage->getRelativePath()) / index_name;
|
||||
String index_name = "primary" + getIndexExtensionFromFilesystem(getDataPartStorage()).value();
|
||||
String index_path = fs::path(getDataPartStorage().getRelativePath()) / index_name;
|
||||
auto index_file = metadata_manager->read(index_name);
|
||||
size_t marks_count = index_granularity.getMarksCount();
|
||||
|
||||
@ -781,7 +784,7 @@ void IMergeTreeDataPart::appendFilesOfIndex(Strings & files) const
|
||||
|
||||
if (metadata_snapshot->hasPrimaryKey())
|
||||
{
|
||||
String index_name = "primary" + getIndexExtensionFromFilesystem(data_part_storage).value();
|
||||
String index_name = "primary" + getIndexExtensionFromFilesystem(getDataPartStorage()).value();
|
||||
files.push_back(index_name);
|
||||
}
|
||||
}
|
||||
@ -793,10 +796,10 @@ NameSet IMergeTreeDataPart::getFileNamesWithoutChecksums() const
|
||||
|
||||
NameSet result = {"checksums.txt", "columns.txt"};
|
||||
|
||||
if (data_part_storage->exists(DEFAULT_COMPRESSION_CODEC_FILE_NAME))
|
||||
if (getDataPartStorage().exists(DEFAULT_COMPRESSION_CODEC_FILE_NAME))
|
||||
result.emplace(DEFAULT_COMPRESSION_CODEC_FILE_NAME);
|
||||
|
||||
if (data_part_storage->exists(TXN_VERSION_METADATA_FILE_NAME))
|
||||
if (getDataPartStorage().exists(TXN_VERSION_METADATA_FILE_NAME))
|
||||
result.emplace(TXN_VERSION_METADATA_FILE_NAME);
|
||||
|
||||
return result;
|
||||
@ -811,7 +814,7 @@ void IMergeTreeDataPart::loadDefaultCompressionCodec()
|
||||
return;
|
||||
}
|
||||
|
||||
String path = fs::path(data_part_storage->getRelativePath()) / DEFAULT_COMPRESSION_CODEC_FILE_NAME;
|
||||
String path = fs::path(getDataPartStorage().getRelativePath()) / DEFAULT_COMPRESSION_CODEC_FILE_NAME;
|
||||
bool exists = metadata_manager->exists(DEFAULT_COMPRESSION_CODEC_FILE_NAME);
|
||||
if (!exists)
|
||||
{
|
||||
@ -851,6 +854,120 @@ void IMergeTreeDataPart::loadDefaultCompressionCodec()
|
||||
}
|
||||
}
|
||||
|
||||
template <typename Writer>
|
||||
void IMergeTreeDataPart::writeMetadata(const String & filename, const WriteSettings & settings, Writer && writer)
|
||||
{
|
||||
auto & data_part_storage = getDataPartStorage();
|
||||
auto tmp_filename = filename + ".tmp";
|
||||
|
||||
try
|
||||
{
|
||||
{
|
||||
auto out = data_part_storage.writeFile(tmp_filename, 4096, settings);
|
||||
writer(*out);
|
||||
out->finalize();
|
||||
}
|
||||
|
||||
data_part_storage.moveFile(tmp_filename, filename);
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
try
|
||||
{
|
||||
if (data_part_storage.exists(tmp_filename))
|
||||
data_part_storage.removeFile(tmp_filename);
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
tryLogCurrentException("DataPartStorageOnDisk");
|
||||
}
|
||||
|
||||
throw;
|
||||
}
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::writeChecksums(const MergeTreeDataPartChecksums & checksums_, const WriteSettings & settings)
|
||||
{
|
||||
writeMetadata("checksums.txt", settings, [&checksums_](auto & buffer)
|
||||
{
|
||||
checksums_.write(buffer);
|
||||
});
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::writeColumns(const NamesAndTypesList & columns_, const WriteSettings & settings)
|
||||
{
|
||||
writeMetadata("columns.txt", settings, [&columns_](auto & buffer)
|
||||
{
|
||||
columns_.writeText(buffer);
|
||||
});
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::writeVersionMetadata(const VersionMetadata & version_, bool fsync_part_dir) const
|
||||
{
|
||||
static constexpr auto filename = "txn_version.txt";
|
||||
static constexpr auto tmp_filename = "txn_version.txt.tmp";
|
||||
auto & data_part_storage = const_cast<IDataPartStorage &>(getDataPartStorage());
|
||||
|
||||
try
|
||||
{
|
||||
{
|
||||
/// TODO IDisk interface does not allow to open file with O_EXCL flag (for DiskLocal),
|
||||
/// so we create empty file at first (expecting that createFile throws if file already exists)
|
||||
/// and then overwrite it.
|
||||
data_part_storage.createFile(tmp_filename);
|
||||
auto write_settings = storage.getContext()->getWriteSettings();
|
||||
auto buf = data_part_storage.writeFile(tmp_filename, 256, write_settings);
|
||||
version_.write(*buf);
|
||||
buf->finalize();
|
||||
buf->sync();
|
||||
}
|
||||
|
||||
SyncGuardPtr sync_guard;
|
||||
if (fsync_part_dir)
|
||||
sync_guard = data_part_storage.getDirectorySyncGuard();
|
||||
data_part_storage.replaceFile(tmp_filename, filename);
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
try
|
||||
{
|
||||
if (data_part_storage.exists(tmp_filename))
|
||||
data_part_storage.removeFile(tmp_filename);
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
tryLogCurrentException("DataPartStorageOnDisk");
|
||||
}
|
||||
|
||||
throw;
|
||||
}
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::writeDeleteOnDestroyMarker()
|
||||
{
|
||||
static constexpr auto marker_path = "delete-on-destroy.txt";
|
||||
|
||||
try
|
||||
{
|
||||
getDataPartStorage().createFile(marker_path);
|
||||
}
|
||||
catch (Poco::Exception & e)
|
||||
{
|
||||
LOG_ERROR(storage.log, "{} (while creating DeleteOnDestroy marker: {})",
|
||||
e.what(), (fs::path(getDataPartStorage().getFullPath()) / marker_path).string());
|
||||
}
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::removeDeleteOnDestroyMarker()
|
||||
{
|
||||
getDataPartStorage().removeFileIfExists("delete-on-destroy.txt");
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::removeVersionMetadata()
|
||||
{
|
||||
getDataPartStorage().removeFileIfExists("txn_version.txt");
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::appendFilesOfDefaultCompressionCodec(Strings & files)
|
||||
{
|
||||
files.push_back(DEFAULT_COMPRESSION_CODEC_FILE_NAME);
|
||||
@ -880,7 +997,7 @@ CompressionCodecPtr IMergeTreeDataPart::detectDefaultCompressionCodec() const
|
||||
String candidate_path = /*fs::path(getRelativePath()) */ (ISerialization::getFileNameForStream(part_column, substream_path) + ".bin");
|
||||
|
||||
/// We can have existing, but empty .bin files. Example: LowCardinality(Nullable(...)) columns and column_name.dict.null.bin file.
|
||||
if (data_part_storage->exists(candidate_path) && data_part_storage->getFileSize(candidate_path) != 0)
|
||||
if (getDataPartStorage().exists(candidate_path) && getDataPartStorage().getFileSize(candidate_path) != 0)
|
||||
path_to_data_file = candidate_path;
|
||||
}
|
||||
});
|
||||
@ -891,7 +1008,7 @@ CompressionCodecPtr IMergeTreeDataPart::detectDefaultCompressionCodec() const
|
||||
continue;
|
||||
}
|
||||
|
||||
result = getCompressionCodecForFile(data_part_storage, path_to_data_file);
|
||||
result = getCompressionCodecForFile(getDataPartStorage(), path_to_data_file);
|
||||
break;
|
||||
}
|
||||
}
|
||||
@ -936,7 +1053,7 @@ void IMergeTreeDataPart::loadPartitionAndMinMaxIndex()
|
||||
String calculated_partition_id = partition.getID(metadata_snapshot->getPartitionKey().sample_block);
|
||||
if (calculated_partition_id != info.partition_id)
|
||||
throw Exception(
|
||||
"While loading part " + data_part_storage->getFullPath() + ": calculated partition ID: " + calculated_partition_id
|
||||
"While loading part " + getDataPartStorage().getFullPath() + ": calculated partition ID: " + calculated_partition_id
|
||||
+ " differs from partition ID in part name: " + info.partition_id,
|
||||
ErrorCodes::CORRUPTED_DATA);
|
||||
}
|
||||
@ -965,7 +1082,7 @@ void IMergeTreeDataPart::loadChecksums(bool require)
|
||||
bytes_on_disk = checksums.getTotalSizeOnDisk();
|
||||
}
|
||||
else
|
||||
bytes_on_disk = data_part_storage->calculateTotalSizeOnDisk();
|
||||
bytes_on_disk = getDataPartStorage().calculateTotalSizeOnDisk();
|
||||
}
|
||||
else
|
||||
{
|
||||
@ -977,7 +1094,7 @@ void IMergeTreeDataPart::loadChecksums(bool require)
|
||||
LOG_WARNING(storage.log, "Checksums for part {} not found. Will calculate them from data on disk.", name);
|
||||
|
||||
checksums = checkDataPart(shared_from_this(), false);
|
||||
data_part_storage->writeChecksums(checksums, {});
|
||||
writeChecksums(checksums, {});
|
||||
|
||||
bytes_on_disk = checksums.getTotalSizeOnDisk();
|
||||
}
|
||||
@ -990,8 +1107,6 @@ void IMergeTreeDataPart::appendFilesOfChecksums(Strings & files)
|
||||
|
||||
void IMergeTreeDataPart::loadRowsCount()
|
||||
{
|
||||
//String path = fs::path(getRelativePath()) / "count.txt";
|
||||
|
||||
auto read_rows_count = [&]()
|
||||
{
|
||||
auto buf = metadata_manager->read("count.txt");
|
||||
@ -1062,7 +1177,7 @@ void IMergeTreeDataPart::loadRowsCount()
|
||||
}
|
||||
else
|
||||
{
|
||||
if (data_part_storage->exists("count.txt"))
|
||||
if (getDataPartStorage().exists("count.txt"))
|
||||
{
|
||||
read_rows_count();
|
||||
return;
|
||||
@ -1161,7 +1276,7 @@ void IMergeTreeDataPart::appendFilesOfUUID(Strings & files)
|
||||
|
||||
void IMergeTreeDataPart::loadColumns(bool require)
|
||||
{
|
||||
String path = fs::path(data_part_storage->getRelativePath()) / "columns.txt";
|
||||
String path = fs::path(getDataPartStorage().getRelativePath()) / "columns.txt";
|
||||
auto metadata_snapshot = storage.getInMemoryMetadataPtr();
|
||||
if (parent_part)
|
||||
metadata_snapshot = metadata_snapshot->projections.get(name).metadata;
|
||||
@ -1172,18 +1287,18 @@ void IMergeTreeDataPart::loadColumns(bool require)
|
||||
{
|
||||
/// We can get list of columns only from columns.txt in compact parts.
|
||||
if (require || part_type == Type::Compact)
|
||||
throw Exception("No columns.txt in part " + name + ", expected path " + path + " on drive " + data_part_storage->getDiskName(),
|
||||
throw Exception("No columns.txt in part " + name + ", expected path " + path + " on drive " + getDataPartStorage().getDiskName(),
|
||||
ErrorCodes::NO_FILE_IN_DATA_PART);
|
||||
|
||||
/// If there is no file with a list of columns, write it down.
|
||||
for (const NameAndTypePair & column : metadata_snapshot->getColumns().getAllPhysical())
|
||||
if (data_part_storage->exists(getFileNameForColumn(column) + ".bin"))
|
||||
if (getDataPartStorage().exists(getFileNameForColumn(column) + ".bin"))
|
||||
loaded_columns.push_back(column);
|
||||
|
||||
if (columns.empty())
|
||||
throw Exception("No columns in part " + name, ErrorCodes::NO_FILE_IN_DATA_PART);
|
||||
|
||||
data_part_storage->writeColumns(loaded_columns, {});
|
||||
writeColumns(loaded_columns, {});
|
||||
}
|
||||
else
|
||||
{
|
||||
@ -1227,7 +1342,7 @@ void IMergeTreeDataPart::assertHasVersionMetadata(MergeTreeTransaction * txn) co
|
||||
name, storage.getStorageID().getNameForLogs(), version.creation_tid, txn ? txn->dumpDescription() : "<none>");
|
||||
|
||||
assert(!txn || storage.supportsTransactions());
|
||||
assert(!txn || data_part_storage->exists(TXN_VERSION_METADATA_FILE_NAME));
|
||||
assert(!txn || getDataPartStorage().exists(TXN_VERSION_METADATA_FILE_NAME));
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::storeVersionMetadata(bool force) const
|
||||
@ -1242,7 +1357,7 @@ void IMergeTreeDataPart::storeVersionMetadata(bool force) const
|
||||
throw Exception(ErrorCodes::NOT_IMPLEMENTED, "Transactions are not supported for in-memory parts (table: {}, part: {})",
|
||||
storage.getStorageID().getNameForLogs(), name);
|
||||
|
||||
data_part_storage->writeVersionMetadata(version, storage.getSettings()->fsync_part_directory);
|
||||
writeVersionMetadata(version, storage.getSettings()->fsync_part_directory);
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::appendCSNToVersionMetadata(VersionMetadata::WhichCSN which_csn) const
|
||||
@ -1254,7 +1369,14 @@ void IMergeTreeDataPart::appendCSNToVersionMetadata(VersionMetadata::WhichCSN wh
|
||||
chassert(!(which_csn == VersionMetadata::WhichCSN::REMOVAL && version.removal_csn == 0));
|
||||
chassert(isStoredOnDisk());
|
||||
|
||||
data_part_storage->appendCSNToVersionMetadata(version, which_csn);
|
||||
/// Small enough appends to file are usually atomic,
|
||||
/// so we append new metadata instead of rewriting file to reduce number of fsyncs.
|
||||
/// We don't need to do fsync when writing CSN, because in case of hard restart
|
||||
/// we will be able to restore CSN from transaction log in Keeper.
|
||||
|
||||
auto out = getDataPartStorage().writeTransactionFile(WriteMode::Append);
|
||||
version.writeCSN(*out, which_csn);
|
||||
out->finalize();
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::appendRemovalTIDToVersionMetadata(bool clear) const
|
||||
@ -1277,13 +1399,74 @@ void IMergeTreeDataPart::appendRemovalTIDToVersionMetadata(bool clear) const
|
||||
else
|
||||
LOG_TEST(storage.log, "Appending removal TID for {} (creation: {}, removal {})", name, version.creation_tid, version.removal_tid);
|
||||
|
||||
data_part_storage->appendRemovalTIDToVersionMetadata(version, clear);
|
||||
auto out = getDataPartStorage().writeTransactionFile(WriteMode::Append);
|
||||
version.writeRemovalTID(*out, clear);
|
||||
out->finalize();
|
||||
|
||||
/// fsync is not required when we clearing removal TID, because after hard restart we will fix metadata
|
||||
if (!clear)
|
||||
out->sync();
|
||||
}
|
||||
|
||||
static std::unique_ptr<ReadBufferFromFileBase> openForReading(const IDataPartStorage & part_storage, const String & filename)
|
||||
{
|
||||
size_t file_size = part_storage.getFileSize(filename);
|
||||
return part_storage.readFile(filename, ReadSettings().adjustBufferSize(file_size), file_size, file_size);
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::loadVersionMetadata() const
|
||||
try
|
||||
{
|
||||
data_part_storage->loadVersionMetadata(version, storage.log);
|
||||
static constexpr auto version_file_name = "txn_version.txt";
|
||||
static constexpr auto tmp_version_file_name = "txn_version.txt.tmp";
|
||||
auto & data_part_storage = const_cast<IDataPartStorage &>(getDataPartStorage());
|
||||
|
||||
auto remove_tmp_file = [&]()
|
||||
{
|
||||
auto last_modified = data_part_storage.getLastModified();
|
||||
auto buf = openForReading(data_part_storage, tmp_version_file_name);
|
||||
|
||||
String content;
|
||||
readStringUntilEOF(content, *buf);
|
||||
LOG_WARNING(storage.log, "Found file {} that was last modified on {}, has size {} and the following content: {}",
|
||||
tmp_version_file_name, last_modified.epochTime(), content.size(), content);
|
||||
data_part_storage.removeFile(tmp_version_file_name);
|
||||
};
|
||||
|
||||
if (data_part_storage.exists(version_file_name))
|
||||
{
|
||||
auto buf = openForReading(data_part_storage, version_file_name);
|
||||
version.read(*buf);
|
||||
if (data_part_storage.exists(tmp_version_file_name))
|
||||
remove_tmp_file();
|
||||
return;
|
||||
}
|
||||
|
||||
/// Four (?) cases are possible:
|
||||
/// 1. Part was created without transactions.
|
||||
/// 2. Version metadata file was not renamed from *.tmp on part creation.
|
||||
/// 3. Version metadata were written to *.tmp file, but hard restart happened before fsync.
|
||||
/// 4. Fsyncs in storeVersionMetadata() work incorrectly.
|
||||
|
||||
if (!data_part_storage.exists(tmp_version_file_name))
|
||||
{
|
||||
/// Case 1.
|
||||
/// We do not have version metadata and transactions history for old parts,
|
||||
/// so let's consider that such parts were created by some ancient transaction
|
||||
/// and were committed with some prehistoric CSN.
|
||||
/// NOTE It might be Case 3, but version metadata file is written on part creation before other files,
|
||||
/// so it's not Case 3 if part is not broken.
|
||||
version.setCreationTID(Tx::PrehistoricTID, nullptr);
|
||||
version.creation_csn = Tx::PrehistoricCSN;
|
||||
return;
|
||||
}
|
||||
|
||||
/// Case 2.
|
||||
/// Content of *.tmp file may be broken, just use fake TID.
|
||||
/// Transaction was not committed if *.tmp file was not renamed, so we should complete rollback by removing part.
|
||||
version.setCreationTID(Tx::DummyTID, nullptr);
|
||||
version.creation_csn = Tx::RolledBackCSN;
|
||||
remove_tmp_file();
|
||||
}
|
||||
catch (Exception & e)
|
||||
{
|
||||
@ -1320,15 +1503,15 @@ bool IMergeTreeDataPart::assertHasValidVersionMetadata() const
|
||||
if (state == MergeTreeDataPartState::Temporary)
|
||||
return true;
|
||||
|
||||
if (!data_part_storage->exists())
|
||||
if (!getDataPartStorage().exists())
|
||||
return true;
|
||||
|
||||
String content;
|
||||
String version_file_name = TXN_VERSION_METADATA_FILE_NAME;
|
||||
try
|
||||
{
|
||||
size_t file_size = data_part_storage->getFileSize(TXN_VERSION_METADATA_FILE_NAME);
|
||||
auto buf = data_part_storage->readFile(TXN_VERSION_METADATA_FILE_NAME, ReadSettings().adjustBufferSize(file_size), file_size, std::nullopt);
|
||||
size_t file_size = getDataPartStorage().getFileSize(TXN_VERSION_METADATA_FILE_NAME);
|
||||
auto buf = getDataPartStorage().readFile(TXN_VERSION_METADATA_FILE_NAME, ReadSettings().adjustBufferSize(file_size), file_size, std::nullopt);
|
||||
|
||||
readStringUntilEOF(content, *buf);
|
||||
ReadBufferFromString str_buf{content};
|
||||
@ -1362,10 +1545,11 @@ void IMergeTreeDataPart::appendFilesOfColumns(Strings & files)
|
||||
|
||||
bool IMergeTreeDataPart::shallParticipateInMerges(const StoragePolicyPtr & storage_policy) const
|
||||
{
|
||||
return data_part_storage->shallParticipateInMerges(*storage_policy);
|
||||
auto disk_name = getDataPartStorage().getDiskName();
|
||||
return !storage_policy->getVolumeByDiskName(disk_name)->areMergesAvoided();
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::renameTo(const String & new_relative_path, bool remove_new_dir_if_exists, DataPartStorageBuilderPtr builder) const
|
||||
void IMergeTreeDataPart::renameTo(const String & new_relative_path, bool remove_new_dir_if_exists)
|
||||
try
|
||||
{
|
||||
assertOnDisk();
|
||||
@ -1376,22 +1560,21 @@ try
|
||||
if (parent_part)
|
||||
{
|
||||
/// For projections, move is only possible inside parent part dir.
|
||||
relative_path = parent_part->data_part_storage->getRelativePath();
|
||||
relative_path = parent_part->getDataPartStorage().getRelativePath();
|
||||
}
|
||||
|
||||
String from = data_part_storage->getRelativePath();
|
||||
auto old_projection_root_path = getDataPartStorage().getRelativePath();
|
||||
auto to = fs::path(relative_path) / new_relative_path;
|
||||
|
||||
metadata_manager->deleteAll(true);
|
||||
metadata_manager->assertAllDeleted(true);
|
||||
builder->rename(to.parent_path(), to.filename(), storage.log, remove_new_dir_if_exists, fsync_dir);
|
||||
data_part_storage->onRename(to.parent_path(), to.filename());
|
||||
getDataPartStorage().rename(to.parent_path(), to.filename(), storage.log, remove_new_dir_if_exists, fsync_dir);
|
||||
metadata_manager->updateAll(true);
|
||||
|
||||
for (const auto & [p_name, part] : projection_parts)
|
||||
{
|
||||
part->data_part_storage = data_part_storage->getProjection(p_name + ".proj");
|
||||
}
|
||||
auto new_projection_root_path = to.string();
|
||||
|
||||
for (const auto & [_, part] : projection_parts)
|
||||
part->getDataPartStorage().changeRootPath(old_projection_root_path, new_projection_root_path);
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
@ -1432,14 +1615,14 @@ void IMergeTreeDataPart::initializePartMetadataManager()
|
||||
|
||||
void IMergeTreeDataPart::initializeIndexGranularityInfo()
|
||||
{
|
||||
auto mrk_ext = MergeTreeIndexGranularityInfo::getMarksExtensionFromFilesystem(data_part_storage);
|
||||
auto mrk_ext = MergeTreeIndexGranularityInfo::getMarksExtensionFromFilesystem(getDataPartStorage());
|
||||
if (mrk_ext)
|
||||
index_granularity_info = MergeTreeIndexGranularityInfo(storage, MarkType{*mrk_ext});
|
||||
else
|
||||
index_granularity_info = MergeTreeIndexGranularityInfo(storage, part_type);
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::remove() const
|
||||
void IMergeTreeDataPart::remove()
|
||||
{
|
||||
assert(assertHasValidVersionMetadata());
|
||||
part_is_probably_removed_from_disk = true;
|
||||
@ -1456,7 +1639,6 @@ void IMergeTreeDataPart::remove() const
|
||||
return CanRemoveDescription{.can_remove_anything = can_remove, .files_not_to_remove = files_not_to_remove };
|
||||
};
|
||||
|
||||
|
||||
if (!isStoredOnDisk())
|
||||
return;
|
||||
|
||||
@ -1475,7 +1657,7 @@ void IMergeTreeDataPart::remove() const
|
||||
projection_checksums.emplace_back(IDataPartStorage::ProjectionChecksums{.name = p_name, .checksums = projection_part->checksums});
|
||||
}
|
||||
|
||||
data_part_storage->remove(std::move(can_remove_callback), checksums, projection_checksums, is_temp, getState(), storage.log);
|
||||
getDataPartStorage().remove(std::move(can_remove_callback), checksums, projection_checksums, is_temp, getState(), storage.log);
|
||||
}
|
||||
|
||||
std::optional<String> IMergeTreeDataPart::getRelativePathForPrefix(const String & prefix, bool detached, bool broken) const
|
||||
@ -1492,7 +1674,7 @@ std::optional<String> IMergeTreeDataPart::getRelativePathForPrefix(const String
|
||||
if (detached && parent_part)
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR, "Cannot detach projection");
|
||||
|
||||
return data_part_storage->getRelativePathForPrefix(storage.log, prefix, detached, broken);
|
||||
return getDataPartStorage().getRelativePathForPrefix(storage.log, prefix, detached, broken);
|
||||
}
|
||||
|
||||
std::optional<String> IMergeTreeDataPart::getRelativePathForDetachedPart(const String & prefix, bool broken) const
|
||||
@ -1507,11 +1689,11 @@ std::optional<String> IMergeTreeDataPart::getRelativePathForDetachedPart(const S
|
||||
return {};
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::renameToDetached(const String & prefix, DataPartStorageBuilderPtr builder) const
|
||||
void IMergeTreeDataPart::renameToDetached(const String & prefix)
|
||||
{
|
||||
auto path_to_detach = getRelativePathForDetachedPart(prefix, /* broken */ false);
|
||||
assert(path_to_detach);
|
||||
renameTo(path_to_detach.value(), true, builder);
|
||||
renameTo(path_to_detach.value(), true);
|
||||
part_is_probably_removed_from_disk = true;
|
||||
}
|
||||
|
||||
@ -1530,7 +1712,7 @@ void IMergeTreeDataPart::makeCloneInDetached(const String & prefix, const Storag
|
||||
if (!maybe_path_in_detached)
|
||||
return;
|
||||
|
||||
data_part_storage->freeze(
|
||||
getDataPartStorage().freeze(
|
||||
storage.relative_data_path,
|
||||
*maybe_path_in_detached,
|
||||
/*make_source_readonly*/ true,
|
||||
@ -1539,17 +1721,17 @@ void IMergeTreeDataPart::makeCloneInDetached(const String & prefix, const Storag
|
||||
{});
|
||||
}
|
||||
|
||||
DataPartStoragePtr IMergeTreeDataPart::makeCloneOnDisk(const DiskPtr & disk, const String & directory_name) const
|
||||
MutableDataPartStoragePtr IMergeTreeDataPart::makeCloneOnDisk(const DiskPtr & disk, const String & directory_name) const
|
||||
{
|
||||
assertOnDisk();
|
||||
|
||||
if (disk->getName() == data_part_storage->getDiskName())
|
||||
throw Exception("Can not clone data part " + name + " to same disk " + data_part_storage->getDiskName(), ErrorCodes::LOGICAL_ERROR);
|
||||
if (disk->getName() == getDataPartStorage().getDiskName())
|
||||
throw Exception("Can not clone data part " + name + " to same disk " + getDataPartStorage().getDiskName(), ErrorCodes::LOGICAL_ERROR);
|
||||
if (directory_name.empty())
|
||||
throw Exception("Can not clone data part " + name + " to empty directory.", ErrorCodes::LOGICAL_ERROR);
|
||||
|
||||
String path_to_clone = fs::path(storage.relative_data_path) / directory_name / "";
|
||||
return data_part_storage->clone(path_to_clone, data_part_storage->getPartDirectory(), disk, storage.log);
|
||||
return getDataPartStorage().clonePart(path_to_clone, getDataPartStorage().getPartDirectory(), disk, storage.log);
|
||||
}
|
||||
|
||||
void IMergeTreeDataPart::checkConsistencyBase() const
|
||||
@ -1590,26 +1772,26 @@ void IMergeTreeDataPart::checkConsistencyBase() const
|
||||
}
|
||||
}
|
||||
|
||||
data_part_storage->checkConsistency(checksums);
|
||||
getDataPartStorage().checkConsistency(checksums);
|
||||
}
|
||||
else
|
||||
{
|
||||
auto check_file_not_empty = [this](const String & file_path)
|
||||
{
|
||||
UInt64 file_size;
|
||||
if (!data_part_storage->exists(file_path) || (file_size = data_part_storage->getFileSize(file_path)) == 0)
|
||||
if (!getDataPartStorage().exists(file_path) || (file_size = getDataPartStorage().getFileSize(file_path)) == 0)
|
||||
throw Exception(
|
||||
ErrorCodes::BAD_SIZE_OF_FILE_IN_DATA_PART,
|
||||
"Part {} is broken: {} is empty",
|
||||
data_part_storage->getFullPath(),
|
||||
std::string(fs::path(data_part_storage->getFullPath()) / file_path));
|
||||
getDataPartStorage().getFullPath(),
|
||||
std::string(fs::path(getDataPartStorage().getFullPath()) / file_path));
|
||||
return file_size;
|
||||
};
|
||||
|
||||
/// Check that the primary key index is not empty.
|
||||
if (!pk.column_names.empty())
|
||||
{
|
||||
String index_name = "primary" + getIndexExtensionFromFilesystem(data_part_storage).value();
|
||||
String index_name = "primary" + getIndexExtensionFromFilesystem(getDataPartStorage()).value();
|
||||
check_file_not_empty(index_name);
|
||||
}
|
||||
|
||||
@ -1753,7 +1935,7 @@ bool IMergeTreeDataPart::checkAllTTLCalculated(const StorageMetadataPtr & metada
|
||||
|
||||
String IMergeTreeDataPart::getUniqueId() const
|
||||
{
|
||||
return data_part_storage->getUniqueId();
|
||||
return getDataPartStorage().getUniqueId();
|
||||
}
|
||||
|
||||
String IMergeTreeDataPart::getZeroLevelPartBlockID(std::string_view token) const
|
||||
@ -1792,11 +1974,11 @@ IMergeTreeDataPart::uint128 IMergeTreeDataPart::getActualChecksumByFile(const St
|
||||
return it->second.file_hash;
|
||||
}
|
||||
|
||||
if (!data_part_storage->exists(file_name))
|
||||
if (!getDataPartStorage().exists(file_name))
|
||||
{
|
||||
return {};
|
||||
}
|
||||
std::unique_ptr<ReadBufferFromFileBase> in_file = data_part_storage->readFile(file_name, {}, std::nullopt, std::nullopt);
|
||||
std::unique_ptr<ReadBufferFromFileBase> in_file = getDataPartStorage().readFile(file_name, {}, std::nullopt, std::nullopt);
|
||||
HashingReadBuffer in_hash(*in_file);
|
||||
|
||||
String value;
|
||||
@ -1824,11 +2006,11 @@ bool isInMemoryPart(const MergeTreeDataPartPtr & data_part)
|
||||
return (data_part && data_part->getType() == MergeTreeDataPartType::InMemory);
|
||||
}
|
||||
|
||||
std::optional<std::string> getIndexExtensionFromFilesystem(const DataPartStoragePtr & data_part_storage)
|
||||
std::optional<std::string> getIndexExtensionFromFilesystem(const IDataPartStorage & data_part_storage)
|
||||
{
|
||||
if (data_part_storage->exists())
|
||||
if (data_part_storage.exists())
|
||||
{
|
||||
for (auto it = data_part_storage->iterate(); it->isValid(); it->next())
|
||||
for (auto it = data_part_storage.iterate(); it->isValid(); it->next())
|
||||
{
|
||||
const auto & extension = fs::path(it->name()).extension();
|
||||
if (extension == getIndexExtension(false)
|
||||
|
||||
@ -1,5 +1,6 @@
|
||||
#pragma once
|
||||
|
||||
#include "IO/WriteSettings.h"
|
||||
#include <Core/Block.h>
|
||||
#include <base/types.h>
|
||||
#include <Core/NamesAndTypes.h>
|
||||
@ -46,7 +47,7 @@ class UncompressedCache;
|
||||
class MergeTreeTransaction;
|
||||
|
||||
/// Description of the data part.
|
||||
class IMergeTreeDataPart : public std::enable_shared_from_this<IMergeTreeDataPart>
|
||||
class IMergeTreeDataPart : public std::enable_shared_from_this<IMergeTreeDataPart>, public DataPartStorageHolder
|
||||
{
|
||||
public:
|
||||
static constexpr auto DATA_FILE_EXTENSION = ".bin";
|
||||
@ -67,19 +68,18 @@ public:
|
||||
|
||||
using uint128 = IPartMetadataManager::uint128;
|
||||
|
||||
|
||||
IMergeTreeDataPart(
|
||||
const MergeTreeData & storage_,
|
||||
const String & name_,
|
||||
const MergeTreePartInfo & info_,
|
||||
const DataPartStoragePtr & data_part_storage_,
|
||||
const MutableDataPartStoragePtr & data_part_storage_,
|
||||
Type part_type_,
|
||||
const IMergeTreeDataPart * parent_part_);
|
||||
|
||||
IMergeTreeDataPart(
|
||||
const MergeTreeData & storage_,
|
||||
const String & name_,
|
||||
const DataPartStoragePtr & data_part_storage_,
|
||||
const MutableDataPartStoragePtr & data_part_storage_,
|
||||
Type part_type_,
|
||||
const IMergeTreeDataPart * parent_part_);
|
||||
|
||||
@ -94,13 +94,12 @@ public:
|
||||
const ReadBufferFromFileBase::ProfileCallback & profile_callback_) const = 0;
|
||||
|
||||
virtual MergeTreeWriterPtr getWriter(
|
||||
DataPartStorageBuilderPtr data_part_storage_builder,
|
||||
const NamesAndTypesList & columns_list,
|
||||
const StorageMetadataPtr & metadata_snapshot,
|
||||
const std::vector<MergeTreeIndexPtr> & indices_to_recalc,
|
||||
const CompressionCodecPtr & default_codec_,
|
||||
const MergeTreeWriterSettings & writer_settings,
|
||||
const MergeTreeIndexGranularity & computed_index_granularity) const = 0;
|
||||
const MergeTreeIndexGranularity & computed_index_granularity) = 0;
|
||||
|
||||
virtual bool isStoredOnDisk() const = 0;
|
||||
|
||||
@ -152,7 +151,7 @@ public:
|
||||
/// Throws an exception if part is not stored in on-disk format.
|
||||
void assertOnDisk() const;
|
||||
|
||||
void remove() const;
|
||||
void remove();
|
||||
|
||||
/// Initialize columns (from columns.txt if exists, or create from column files if not).
|
||||
/// Load checksums from checksums.txt if exists. Load index if required.
|
||||
@ -200,10 +199,6 @@ public:
|
||||
/// processed by multiple shards.
|
||||
UUID uuid = UUIDHelpers::Nil;
|
||||
|
||||
/// This is an object which encapsulates all the operations with disk.
|
||||
/// Contains a path to stored data.
|
||||
DataPartStoragePtr data_part_storage;
|
||||
|
||||
MergeTreeIndexGranularityInfo index_granularity_info;
|
||||
|
||||
size_t rows_count = 0;
|
||||
@ -289,8 +284,8 @@ public:
|
||||
|
||||
using WrittenFiles = std::vector<std::unique_ptr<WriteBufferFromFileBase>>;
|
||||
|
||||
[[nodiscard]] WrittenFiles store(const MergeTreeData & data, const DataPartStorageBuilderPtr & data_part_storage_builder, Checksums & checksums) const;
|
||||
[[nodiscard]] WrittenFiles store(const Names & column_names, const DataTypes & data_types, const DataPartStorageBuilderPtr & data_part_storage_builder, Checksums & checksums) const;
|
||||
[[nodiscard]] WrittenFiles store(const MergeTreeData & data, IDataPartStorage & part_storage, Checksums & checksums) const;
|
||||
[[nodiscard]] WrittenFiles store(const Names & column_names, const DataTypes & data_types, IDataPartStorage & part_storage, Checksums & checksums) const;
|
||||
|
||||
void update(const Block & block, const Names & column_names);
|
||||
void merge(const MinMaxIndex & other);
|
||||
@ -321,17 +316,17 @@ public:
|
||||
size_t getFileSizeOrZero(const String & file_name) const;
|
||||
|
||||
/// Moves a part to detached/ directory and adds prefix to its name
|
||||
void renameToDetached(const String & prefix, DataPartStorageBuilderPtr builder) const;
|
||||
void renameToDetached(const String & prefix);
|
||||
|
||||
/// Makes checks and move part to new directory
|
||||
/// Changes only relative_dir_name, you need to update other metadata (name, is_temp) explicitly
|
||||
virtual void renameTo(const String & new_relative_path, bool remove_new_dir_if_exists, DataPartStorageBuilderPtr builder) const;
|
||||
virtual void renameTo(const String & new_relative_path, bool remove_new_dir_if_exists);
|
||||
|
||||
/// Makes clone of a part in detached/ directory via hard links
|
||||
virtual void makeCloneInDetached(const String & prefix, const StorageMetadataPtr & metadata_snapshot) const;
|
||||
|
||||
/// Makes full clone of part in specified subdirectory (relative to storage data directory, e.g. "detached") on another disk
|
||||
DataPartStoragePtr makeCloneOnDisk(const DiskPtr & disk, const String & directory_name) const;
|
||||
MutableDataPartStoragePtr makeCloneOnDisk(const DiskPtr & disk, const String & directory_name) const;
|
||||
|
||||
/// Checks that .bin and .mrk files exist.
|
||||
///
|
||||
@ -445,6 +440,12 @@ public:
|
||||
/// True if here is lightweight deleted mask file in part.
|
||||
bool hasLightweightDelete() const { return columns.contains(LightweightDeleteDescription::FILTER_COLUMN.name); }
|
||||
|
||||
void writeChecksums(const MergeTreeDataPartChecksums & checksums_, const WriteSettings & settings);
|
||||
|
||||
void writeDeleteOnDestroyMarker();
|
||||
void removeDeleteOnDestroyMarker();
|
||||
void removeVersionMetadata();
|
||||
|
||||
protected:
|
||||
|
||||
/// Total size of all columns, calculated once in calcuateColumnSizesOnDisk
|
||||
@ -566,6 +567,12 @@ private:
|
||||
/// any specifial compression.
|
||||
void loadDefaultCompressionCodec();
|
||||
|
||||
void writeColumns(const NamesAndTypesList & columns_, const WriteSettings & settings);
|
||||
void writeVersionMetadata(const VersionMetadata & version_, bool fsync_part_dir) const;
|
||||
|
||||
template <typename Writer>
|
||||
void writeMetadata(const String & filename, const WriteSettings & settings, Writer && writer);
|
||||
|
||||
static void appendFilesOfDefaultCompressionCodec(Strings & files);
|
||||
|
||||
/// Found column without specific compression and return codec
|
||||
@ -585,7 +592,7 @@ bool isCompactPart(const MergeTreeDataPartPtr & data_part);
|
||||
bool isWidePart(const MergeTreeDataPartPtr & data_part);
|
||||
bool isInMemoryPart(const MergeTreeDataPartPtr & data_part);
|
||||
inline String getIndexExtension(bool is_compressed_primary_key) { return is_compressed_primary_key ? ".cidx" : ".idx"; }
|
||||
std::optional<String> getIndexExtensionFromFilesystem(const DataPartStoragePtr & data_part_storage);
|
||||
std::optional<String> getIndexExtensionFromFilesystem(const IDataPartStorage & data_part_storage);
|
||||
bool isCompressedFromIndexExtension(const String & index_extension);
|
||||
|
||||
}
|
||||
|
||||
@ -7,7 +7,8 @@ namespace DB
|
||||
{
|
||||
|
||||
class IDataPartStorage;
|
||||
using DataPartStoragePtr = std::shared_ptr<IDataPartStorage>;
|
||||
using DataPartStoragePtr = std::shared_ptr<const IDataPartStorage>;
|
||||
|
||||
class MergeTreeIndexGranularity;
|
||||
struct MergeTreeDataPartChecksums;
|
||||
struct MergeTreeIndexGranularityInfo;
|
||||
@ -36,7 +37,7 @@ public:
|
||||
|
||||
virtual bool isProjectionPart() const = 0;
|
||||
|
||||
virtual const DataPartStoragePtr & getDataPartStorage() const = 0;
|
||||
virtual DataPartStoragePtr getDataPartStorage() const = 0;
|
||||
|
||||
virtual const NamesAndTypesList & getColumns() const = 0;
|
||||
|
||||
|
||||
@ -38,14 +38,12 @@ Block permuteBlockIfNeeded(const Block & block, const IColumn::Permutation * per
|
||||
}
|
||||
|
||||
IMergeTreeDataPartWriter::IMergeTreeDataPartWriter(
|
||||
const MergeTreeData::DataPartPtr & data_part_,
|
||||
DataPartStorageBuilderPtr data_part_storage_builder_,
|
||||
const MergeTreeMutableDataPartPtr & data_part_,
|
||||
const NamesAndTypesList & columns_list_,
|
||||
const StorageMetadataPtr & metadata_snapshot_,
|
||||
const MergeTreeWriterSettings & settings_,
|
||||
const MergeTreeIndexGranularity & index_granularity_)
|
||||
: data_part(data_part_)
|
||||
, data_part_storage_builder(std::move(data_part_storage_builder_))
|
||||
, storage(data_part_->storage)
|
||||
, metadata_snapshot(metadata_snapshot_)
|
||||
, columns_list(columns_list_)
|
||||
|
||||
@ -22,8 +22,7 @@ class IMergeTreeDataPartWriter : private boost::noncopyable
|
||||
{
|
||||
public:
|
||||
IMergeTreeDataPartWriter(
|
||||
const MergeTreeData::DataPartPtr & data_part_,
|
||||
DataPartStorageBuilderPtr data_part_storage_builder_,
|
||||
const MergeTreeMutableDataPartPtr & data_part_,
|
||||
const NamesAndTypesList & columns_list_,
|
||||
const StorageMetadataPtr & metadata_snapshot_,
|
||||
const MergeTreeWriterSettings & settings_,
|
||||
@ -42,8 +41,7 @@ public:
|
||||
|
||||
protected:
|
||||
|
||||
const MergeTreeData::DataPartPtr data_part;
|
||||
DataPartStorageBuilderPtr data_part_storage_builder;
|
||||
const MergeTreeMutableDataPartPtr data_part;
|
||||
const MergeTreeData & storage;
|
||||
const StorageMetadataPtr metadata_snapshot;
|
||||
const NamesAndTypesList columns_list;
|
||||
|
||||
@ -6,14 +6,13 @@ namespace DB
|
||||
{
|
||||
|
||||
IMergedBlockOutputStream::IMergedBlockOutputStream(
|
||||
DataPartStorageBuilderPtr data_part_storage_builder_,
|
||||
const MergeTreeDataPartPtr & data_part,
|
||||
const MergeTreeMutableDataPartPtr & data_part,
|
||||
const StorageMetadataPtr & metadata_snapshot_,
|
||||
const NamesAndTypesList & columns_list,
|
||||
bool reset_columns_)
|
||||
: storage(data_part->storage)
|
||||
, metadata_snapshot(metadata_snapshot_)
|
||||
, data_part_storage_builder(std::move(data_part_storage_builder_))
|
||||
, data_part_storage(data_part->getDataPartStoragePtr())
|
||||
, reset_columns(reset_columns_)
|
||||
{
|
||||
if (reset_columns)
|
||||
|
||||
@ -1,5 +1,6 @@
|
||||
#pragma once
|
||||
|
||||
#include "Storages/MergeTree/IDataPartStorage.h"
|
||||
#include <Storages/MergeTree/MergeTreeIndexGranularity.h>
|
||||
#include <Storages/MergeTree/MergeTreeData.h>
|
||||
#include <Storages/MergeTree/IMergeTreeDataPart.h>
|
||||
@ -12,8 +13,7 @@ class IMergedBlockOutputStream
|
||||
{
|
||||
public:
|
||||
IMergedBlockOutputStream(
|
||||
DataPartStorageBuilderPtr data_part_storage_builder_,
|
||||
const MergeTreeDataPartPtr & data_part,
|
||||
const MergeTreeMutableDataPartPtr & data_part,
|
||||
const StorageMetadataPtr & metadata_snapshot_,
|
||||
const NamesAndTypesList & columns_list,
|
||||
bool reset_columns_);
|
||||
@ -42,7 +42,7 @@ protected:
|
||||
const MergeTreeData & storage;
|
||||
StorageMetadataPtr metadata_snapshot;
|
||||
|
||||
DataPartStorageBuilderPtr data_part_storage_builder;
|
||||
MutableDataPartStoragePtr data_part_storage;
|
||||
IMergeTreeDataPart::MergeTreeWriterPtr writer;
|
||||
|
||||
bool reset_columns = false;
|
||||
|
||||
@ -1424,6 +1424,7 @@ public:
|
||||
ColumnsWithTypeAndName new_arguments;
|
||||
new_arguments.reserve(arguments.size() + 1);
|
||||
new_arguments.push_back(const_arg);
|
||||
new_arguments.front().column = new_arguments.front().column->cloneResized(input_rows_count);
|
||||
for (const auto & arg : arguments)
|
||||
new_arguments.push_back(arg);
|
||||
return func->prepare(new_arguments)->execute(new_arguments, result_type, input_rows_count, dry_run);
|
||||
@ -1432,6 +1433,7 @@ public:
|
||||
{
|
||||
auto new_arguments = arguments;
|
||||
new_arguments.push_back(const_arg);
|
||||
new_arguments.back().column = new_arguments.back().column->cloneResized(input_rows_count);
|
||||
return func->prepare(new_arguments)->execute(new_arguments, result_type, input_rows_count, dry_run);
|
||||
}
|
||||
else
|
||||
|
||||
@ -12,7 +12,8 @@ public:
|
||||
explicit LoadedMergeTreeDataPartInfoForReader(MergeTreeData::DataPartPtr data_part_)
|
||||
: IMergeTreeDataPartInfoForReader(data_part_->storage.getContext())
|
||||
, data_part(data_part_)
|
||||
{}
|
||||
{
|
||||
}
|
||||
|
||||
bool isCompactPart() const override { return DB::isCompactPart(data_part); }
|
||||
|
||||
@ -22,7 +23,7 @@ public:
|
||||
|
||||
bool isProjectionPart() const override { return data_part->isProjectionPart(); }
|
||||
|
||||
const DataPartStoragePtr & getDataPartStorage() const override { return data_part->data_part_storage; }
|
||||
DataPartStoragePtr getDataPartStorage() const override { return data_part->getDataPartStoragePtr(); }
|
||||
|
||||
const NamesAndTypesList & getColumns() const override { return data_part->getColumns(); }
|
||||
|
||||
|
||||
@ -160,7 +160,9 @@ ReplicatedMergeMutateTaskBase::PrepareResult MergeFromLogEntryTask::prepare()
|
||||
for (auto & part_ptr : parts)
|
||||
{
|
||||
ttl_infos.update(part_ptr->ttl_infos);
|
||||
max_volume_index = std::max(max_volume_index, part_ptr->data_part_storage->getVolumeIndex(*storage.getStoragePolicy()));
|
||||
auto disk_name = part_ptr->getDataPartStorage().getDiskName();
|
||||
size_t volume_index = storage.getStoragePolicy()->getVolumeIndexByDiskName(disk_name);
|
||||
max_volume_index = std::max(max_volume_index, volume_index);
|
||||
}
|
||||
|
||||
/// It will live until the whole task is being destroyed
|
||||
@ -294,12 +296,10 @@ ReplicatedMergeMutateTaskBase::PrepareResult MergeFromLogEntryTask::prepare()
|
||||
bool MergeFromLogEntryTask::finalize(ReplicatedMergeMutateTaskBase::PartLogWriter write_part_log)
|
||||
{
|
||||
part = merge_task->getFuture().get();
|
||||
auto builder = merge_task->getBuilder();
|
||||
|
||||
/// Task is not needed
|
||||
merge_task.reset();
|
||||
|
||||
storage.merger_mutator.renameMergedTemporaryPart(part, parts, NO_TRANSACTION_PTR, *transaction_ptr, builder);
|
||||
storage.merger_mutator.renameMergedTemporaryPart(part, parts, NO_TRANSACTION_PTR, *transaction_ptr);
|
||||
|
||||
try
|
||||
{
|
||||
|
||||
@ -65,7 +65,7 @@ MergeListElement::MergeListElement(
|
||||
for (const auto & source_part : future_part->parts)
|
||||
{
|
||||
source_part_names.emplace_back(source_part->name);
|
||||
source_part_paths.emplace_back(source_part->data_part_storage->getFullPath());
|
||||
source_part_paths.emplace_back(source_part->getDataPartStorage().getFullPath());
|
||||
|
||||
total_size_bytes_compressed += source_part->getBytesOnDisk();
|
||||
total_size_marks += source_part->getMarksCount();
|
||||
|
||||
@ -115,10 +115,9 @@ void MergePlainMergeTreeTask::prepare()
|
||||
void MergePlainMergeTreeTask::finish()
|
||||
{
|
||||
new_part = merge_task->getFuture().get();
|
||||
auto builder = merge_task->getBuilder();
|
||||
|
||||
MergeTreeData::Transaction transaction(storage, txn.get());
|
||||
storage.merger_mutator.renameMergedTemporaryPart(new_part, future_part->parts, txn, transaction, builder);
|
||||
storage.merger_mutator.renameMergedTemporaryPart(new_part, future_part->parts, txn, transaction);
|
||||
transaction.commit();
|
||||
|
||||
write_part_log({});
|
||||
|
||||
@ -1,3 +1,4 @@
|
||||
#include "Storages/MergeTree/IDataPartStorage.h"
|
||||
#include <Storages/MergeTree/MergeTask.h>
|
||||
|
||||
#include <memory>
|
||||
@ -125,23 +126,26 @@ bool MergeTask::ExecuteAndFinalizeHorizontalPart::prepare()
|
||||
ctx->disk = global_ctx->space_reservation->getDisk();
|
||||
|
||||
String local_tmp_part_basename = local_tmp_prefix + global_ctx->future_part->name + local_tmp_suffix;
|
||||
MutableDataPartStoragePtr data_part_storage;
|
||||
|
||||
if (global_ctx->parent_path_storage_builder)
|
||||
if (global_ctx->parent_part)
|
||||
{
|
||||
global_ctx->data_part_storage_builder = global_ctx->parent_path_storage_builder->getProjection(local_tmp_part_basename);
|
||||
data_part_storage = global_ctx->parent_part->getDataPartStorage().getProjection(local_tmp_part_basename);
|
||||
}
|
||||
else
|
||||
{
|
||||
auto local_single_disk_volume = std::make_shared<SingleDiskVolume>("volume_" + global_ctx->future_part->name, ctx->disk, 0);
|
||||
|
||||
global_ctx->data_part_storage_builder = std::make_shared<DataPartStorageBuilderOnDisk>(
|
||||
data_part_storage = std::make_shared<DataPartStorageOnDisk>(
|
||||
local_single_disk_volume,
|
||||
global_ctx->data->relative_data_path,
|
||||
local_tmp_part_basename);
|
||||
|
||||
data_part_storage->beginTransaction();
|
||||
}
|
||||
|
||||
if (global_ctx->data_part_storage_builder->exists())
|
||||
throw Exception("Directory " + global_ctx->data_part_storage_builder->getFullPath() + " already exists", ErrorCodes::DIRECTORY_ALREADY_EXISTS);
|
||||
if (data_part_storage->exists())
|
||||
throw Exception("Directory " + data_part_storage->getFullPath() + " already exists", ErrorCodes::DIRECTORY_ALREADY_EXISTS);
|
||||
|
||||
if (!global_ctx->parent_part)
|
||||
global_ctx->temporary_directory_lock = global_ctx->data->getTemporaryPartDirectoryHolder(local_tmp_part_basename);
|
||||
@ -163,8 +167,6 @@ bool MergeTask::ExecuteAndFinalizeHorizontalPart::prepare()
|
||||
global_ctx->merging_columns,
|
||||
global_ctx->merging_column_names);
|
||||
|
||||
auto data_part_storage = global_ctx->data_part_storage_builder->getStorage();
|
||||
|
||||
global_ctx->new_data_part = global_ctx->data->createPart(
|
||||
global_ctx->future_part->name,
|
||||
global_ctx->future_part->type,
|
||||
@ -302,7 +304,6 @@ bool MergeTask::ExecuteAndFinalizeHorizontalPart::prepare()
|
||||
|
||||
global_ctx->to = std::make_shared<MergedBlockOutputStream>(
|
||||
global_ctx->new_data_part,
|
||||
global_ctx->data_part_storage_builder,
|
||||
global_ctx->metadata_snapshot,
|
||||
global_ctx->merging_columns,
|
||||
MergeTreeIndexFactory::instance().getMany(global_ctx->metadata_snapshot->getSecondaryIndices()),
|
||||
@ -501,7 +502,6 @@ void MergeTask::VerticalMergeStage::prepareVerticalMergeForOneColumn() const
|
||||
ctx->executor = std::make_unique<PullingPipelineExecutor>(ctx->column_parts_pipeline);
|
||||
|
||||
ctx->column_to = std::make_unique<MergedColumnOnlyOutputStream>(
|
||||
global_ctx->data_part_storage_builder,
|
||||
global_ctx->new_data_part,
|
||||
global_ctx->metadata_snapshot,
|
||||
ctx->executor->getHeader(),
|
||||
@ -654,7 +654,6 @@ bool MergeTask::MergeProjectionsStage::mergeMinMaxIndexAndPrepareProjections() c
|
||||
global_ctx->deduplicate_by_columns,
|
||||
projection_merging_params,
|
||||
global_ctx->new_data_part.get(),
|
||||
global_ctx->data_part_storage_builder.get(),
|
||||
".proj",
|
||||
NO_TRANSACTION_PTR,
|
||||
global_ctx->data,
|
||||
|
||||
@ -59,8 +59,7 @@ public:
|
||||
bool deduplicate_,
|
||||
Names deduplicate_by_columns_,
|
||||
MergeTreeData::MergingParams merging_params_,
|
||||
const IMergeTreeDataPart * parent_part_,
|
||||
const IDataPartStorageBuilder * parent_path_storage_builder_,
|
||||
IMergeTreeDataPart * parent_part_,
|
||||
String suffix_,
|
||||
MergeTreeTransactionPtr txn,
|
||||
MergeTreeData * data_,
|
||||
@ -82,7 +81,6 @@ public:
|
||||
global_ctx->deduplicate = std::move(deduplicate_);
|
||||
global_ctx->deduplicate_by_columns = std::move(deduplicate_by_columns_);
|
||||
global_ctx->parent_part = std::move(parent_part_);
|
||||
global_ctx->parent_path_storage_builder = std::move(parent_path_storage_builder_);
|
||||
global_ctx->data = std::move(data_);
|
||||
global_ctx->mutator = std::move(mutator_);
|
||||
global_ctx->merges_blocker = std::move(merges_blocker_);
|
||||
@ -102,11 +100,6 @@ public:
|
||||
return global_ctx->promise.get_future();
|
||||
}
|
||||
|
||||
DataPartStorageBuilderPtr getBuilder()
|
||||
{
|
||||
return global_ctx->data_part_storage_builder;
|
||||
}
|
||||
|
||||
bool execute();
|
||||
|
||||
private:
|
||||
@ -141,8 +134,7 @@ private:
|
||||
StorageMetadataPtr metadata_snapshot{nullptr};
|
||||
FutureMergedMutatedPartPtr future_part{nullptr};
|
||||
/// This will be either nullptr or new_data_part, so raw pointer is ok.
|
||||
const IMergeTreeDataPart * parent_part{nullptr};
|
||||
const IDataPartStorageBuilder * parent_path_storage_builder{nullptr};
|
||||
IMergeTreeDataPart * parent_part{nullptr};
|
||||
ContextPtr context{nullptr};
|
||||
time_t time_of_merge{0};
|
||||
ReservationSharedPtr space_reservation{nullptr};
|
||||
@ -168,7 +160,6 @@ private:
|
||||
std::unique_ptr<PullingPipelineExecutor> merging_executor;
|
||||
|
||||
MergeTreeData::MutableDataPartPtr new_data_part{nullptr};
|
||||
DataPartStorageBuilderPtr data_part_storage_builder;
|
||||
|
||||
/// If lightweight delete mask is present then some input rows are filtered out right after reading.
|
||||
std::shared_ptr<std::atomic<size_t>> input_rows_filtered{std::make_shared<std::atomic<size_t>>(0)};
|
||||
|
||||
@ -607,7 +607,7 @@ Block MergeTreeBaseSelectProcessor::transformHeader(
|
||||
if (!row_level_column.type->canBeUsedInBooleanContext())
|
||||
{
|
||||
throw Exception("Invalid type for filter in PREWHERE: " + row_level_column.type->getName(),
|
||||
ErrorCodes::LOGICAL_ERROR);
|
||||
ErrorCodes::ILLEGAL_TYPE_OF_COLUMN_FOR_FILTER);
|
||||
}
|
||||
|
||||
block.erase(prewhere_info->row_level_column_name);
|
||||
@ -620,7 +620,7 @@ Block MergeTreeBaseSelectProcessor::transformHeader(
|
||||
if (!prewhere_column.type->canBeUsedInBooleanContext())
|
||||
{
|
||||
throw Exception("Invalid type for filter in PREWHERE: " + prewhere_column.type->getName(),
|
||||
ErrorCodes::LOGICAL_ERROR);
|
||||
ErrorCodes::ILLEGAL_TYPE_OF_COLUMN_FOR_FILTER);
|
||||
}
|
||||
|
||||
if (prewhere_info->remove_prewhere_column)
|
||||
@ -628,13 +628,13 @@ Block MergeTreeBaseSelectProcessor::transformHeader(
|
||||
else
|
||||
{
|
||||
WhichDataType which(removeNullable(recursiveRemoveLowCardinality(prewhere_column.type)));
|
||||
if (which.isInt() || which.isUInt())
|
||||
if (which.isNativeInt() || which.isNativeUInt())
|
||||
prewhere_column.column = prewhere_column.type->createColumnConst(block.rows(), 1u)->convertToFullColumnIfConst();
|
||||
else if (which.isFloat())
|
||||
prewhere_column.column = prewhere_column.type->createColumnConst(block.rows(), 1.0f)->convertToFullColumnIfConst();
|
||||
else
|
||||
throw Exception("Illegal type " + prewhere_column.type->getName() + " of column for filter.",
|
||||
ErrorCodes::ILLEGAL_TYPE_OF_COLUMN_FOR_FILTER);
|
||||
throw Exception(
|
||||
ErrorCodes::ILLEGAL_TYPE_OF_COLUMN_FOR_FILTER, "Illegal type {} of column for filter", prewhere_column.type->getName());
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
@ -943,8 +943,8 @@ Int64 MergeTreeData::getMaxBlockNumber() const
|
||||
}
|
||||
|
||||
void MergeTreeData::loadDataPartsFromDisk(
|
||||
DataPartsVector & broken_parts_to_detach,
|
||||
DataPartsVector & duplicate_parts_to_remove,
|
||||
MutableDataPartsVector & broken_parts_to_detach,
|
||||
MutableDataPartsVector & duplicate_parts_to_remove,
|
||||
ThreadPool & pool,
|
||||
size_t num_parts,
|
||||
std::queue<std::vector<std::pair<String, DiskPtr>>> & parts_queue,
|
||||
@ -1082,7 +1082,6 @@ void MergeTreeData::loadDataPartsFromDisk(
|
||||
if (size_of_part.has_value())
|
||||
part_size_str = formatReadableSizeWithBinarySuffix(*size_of_part);
|
||||
|
||||
|
||||
LOG_ERROR(log,
|
||||
"Detaching broken part {}{} (size: {}). "
|
||||
"If it happened after update, it is likely because of backward incompatibility. "
|
||||
@ -1200,8 +1199,7 @@ void MergeTreeData::loadDataPartsFromDisk(
|
||||
|
||||
|
||||
void MergeTreeData::loadDataPartsFromWAL(
|
||||
DataPartsVector & /* broken_parts_to_detach */,
|
||||
DataPartsVector & duplicate_parts_to_remove,
|
||||
MutableDataPartsVector & duplicate_parts_to_remove,
|
||||
MutableDataPartsVector & parts_from_wal)
|
||||
{
|
||||
for (auto & part : parts_from_wal)
|
||||
@ -1215,7 +1213,7 @@ void MergeTreeData::loadDataPartsFromWAL(
|
||||
{
|
||||
if ((*it)->checksums.getTotalChecksumHex() == part->checksums.getTotalChecksumHex())
|
||||
{
|
||||
LOG_ERROR(log, "Remove duplicate part {}", part->data_part_storage->getFullPath());
|
||||
LOG_ERROR(log, "Remove duplicate part {}", part->getDataPartStorage().getFullPath());
|
||||
duplicate_parts_to_remove.push_back(part);
|
||||
}
|
||||
else
|
||||
@ -1329,8 +1327,8 @@ void MergeTreeData::loadDataParts(bool skip_sanity_checks)
|
||||
auto part_lock = lockParts();
|
||||
data_parts_indexes.clear();
|
||||
|
||||
DataPartsVector broken_parts_to_detach;
|
||||
DataPartsVector duplicate_parts_to_remove;
|
||||
MutableDataPartsVector broken_parts_to_detach;
|
||||
MutableDataPartsVector duplicate_parts_to_remove;
|
||||
|
||||
if (num_parts > 0)
|
||||
loadDataPartsFromDisk(
|
||||
@ -1384,7 +1382,7 @@ void MergeTreeData::loadDataParts(bool skip_sanity_checks)
|
||||
parts_from_wal.insert(
|
||||
parts_from_wal.end(), std::make_move_iterator(disk_wal_parts.begin()), std::make_move_iterator(disk_wal_parts.end()));
|
||||
|
||||
loadDataPartsFromWAL(broken_parts_to_detach, duplicate_parts_to_remove, parts_from_wal);
|
||||
loadDataPartsFromWAL(duplicate_parts_to_remove, parts_from_wal);
|
||||
|
||||
num_parts += parts_from_wal.size();
|
||||
}
|
||||
@ -1397,11 +1395,7 @@ void MergeTreeData::loadDataParts(bool skip_sanity_checks)
|
||||
}
|
||||
|
||||
for (auto & part : broken_parts_to_detach)
|
||||
{
|
||||
auto builder = part->data_part_storage->getBuilder();
|
||||
part->renameToDetached("broken-on-start", builder); /// detached parts must not have '_' in prefixes
|
||||
builder->commit();
|
||||
}
|
||||
part->renameToDetached("broken-on-start"); /// detached parts must not have '_' in prefixes
|
||||
|
||||
for (auto & part : duplicate_parts_to_remove)
|
||||
part->remove();
|
||||
@ -1689,6 +1683,15 @@ scope_guard MergeTreeData::getTemporaryPartDirectoryHolder(const String & part_d
|
||||
return [this, part_dir_name]() { temporary_parts.remove(part_dir_name); };
|
||||
}
|
||||
|
||||
MergeTreeData::MutableDataPartPtr MergeTreeData::preparePartForRemoval(const DataPartPtr & part)
|
||||
{
|
||||
auto state = part->getState();
|
||||
if (state != DataPartState::Deleting && state != DataPartState::DeleteOnDestroy)
|
||||
throw Exception(ErrorCodes::LOGICAL_ERROR,
|
||||
"Cannot remove part {}, because it has state: {}", part->name, magic_enum::enum_name(part->getState()));
|
||||
|
||||
return std::const_pointer_cast<IMergeTreeDataPart>(part);
|
||||
}
|
||||
|
||||
MergeTreeData::DataPartsVector MergeTreeData::grabOldParts(bool force)
|
||||
{
|
||||
@ -1864,7 +1867,7 @@ void MergeTreeData::flushAllInMemoryPartsIfNeeded()
|
||||
{
|
||||
if (auto part_in_memory = asInMemoryPart(part))
|
||||
{
|
||||
part_in_memory->flushToDisk(part_in_memory->data_part_storage->getPartDirectory(), metadata_snapshot);
|
||||
part_in_memory->flushToDisk(part_in_memory->getDataPartStorage().getPartDirectory(), metadata_snapshot);
|
||||
}
|
||||
}
|
||||
}
|
||||
@ -1948,7 +1951,7 @@ void MergeTreeData::clearPartsFromFilesystemImpl(const DataPartsVector & parts_t
|
||||
if (thread_group)
|
||||
CurrentThread::attachToIfDetached(thread_group);
|
||||
|
||||
part->remove();
|
||||
preparePartForRemoval(part)->remove();
|
||||
if (part_names_succeed)
|
||||
{
|
||||
std::lock_guard lock(part_names_mutex);
|
||||
@ -1964,7 +1967,7 @@ void MergeTreeData::clearPartsFromFilesystemImpl(const DataPartsVector & parts_t
|
||||
LOG_DEBUG(log, "Removing {} parts from filesystem: {}", parts_to_remove.size(), fmt::join(parts_to_remove, ", "));
|
||||
for (const DataPartPtr & part : parts_to_remove)
|
||||
{
|
||||
part->remove();
|
||||
preparePartForRemoval(part)->remove();
|
||||
if (part_names_succeed)
|
||||
part_names_succeed->insert(part->name);
|
||||
}
|
||||
@ -2144,11 +2147,14 @@ void MergeTreeData::rename(const String & new_table_path, const StorageID & new_
|
||||
if (!getStorageID().hasUUID())
|
||||
getContext()->dropCaches();
|
||||
|
||||
/// TODO: remove const_cast
|
||||
for (const auto & part : data_parts_by_info)
|
||||
part->data_part_storage->changeRootPath(relative_data_path, new_table_path);
|
||||
{
|
||||
auto & part_mutable = const_cast<IMergeTreeDataPart &>(*part);
|
||||
part_mutable.getDataPartStorage().changeRootPath(relative_data_path, new_table_path);
|
||||
}
|
||||
|
||||
relative_data_path = new_table_path;
|
||||
|
||||
renameInMemory(new_table_id);
|
||||
}
|
||||
|
||||
@ -2166,7 +2172,12 @@ void MergeTreeData::dropAllData()
|
||||
|
||||
auto lock = lockParts();
|
||||
|
||||
DataPartsVector all_parts(data_parts_by_info.begin(), data_parts_by_info.end());
|
||||
DataPartsVector all_parts;
|
||||
for (auto it = data_parts_by_info.begin(); it != data_parts_by_info.end(); ++it)
|
||||
{
|
||||
modifyPartState(it, DataPartState::Deleting);
|
||||
all_parts.push_back(*it);
|
||||
}
|
||||
|
||||
{
|
||||
std::lock_guard wal_lock(write_ahead_log_mutex);
|
||||
@ -2179,7 +2190,6 @@ void MergeTreeData::dropAllData()
|
||||
if (!getStorageID().hasUUID())
|
||||
getContext()->dropCaches();
|
||||
|
||||
|
||||
/// Removing of each data part before recursive removal of directory is to speed-up removal, because there will be less number of syscalls.
|
||||
NameSet part_names_failed;
|
||||
try
|
||||
@ -2726,7 +2736,7 @@ MergeTreeDataPartType MergeTreeData::choosePartTypeOnDisk(size_t bytes_uncompres
|
||||
|
||||
MergeTreeData::MutableDataPartPtr MergeTreeData::createPart(const String & name,
|
||||
MergeTreeDataPartType type, const MergeTreePartInfo & part_info,
|
||||
const DataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part) const
|
||||
const MutableDataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part) const
|
||||
{
|
||||
if (type == MergeTreeDataPartType::Compact)
|
||||
return std::make_shared<MergeTreeDataPartCompact>(*this, name, part_info, data_part_storage, parent_part);
|
||||
@ -2739,17 +2749,17 @@ MergeTreeData::MutableDataPartPtr MergeTreeData::createPart(const String & name,
|
||||
}
|
||||
|
||||
MergeTreeData::MutableDataPartPtr MergeTreeData::createPart(
|
||||
const String & name, const DataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part) const
|
||||
const String & name, const MutableDataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part) const
|
||||
{
|
||||
return createPart(name, MergeTreePartInfo::fromPartName(name, format_version), data_part_storage, parent_part);
|
||||
}
|
||||
|
||||
MergeTreeData::MutableDataPartPtr MergeTreeData::createPart(
|
||||
const String & name, const MergeTreePartInfo & part_info,
|
||||
const DataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part) const
|
||||
const MutableDataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part) const
|
||||
{
|
||||
MergeTreeDataPartType type;
|
||||
auto mrk_ext = MergeTreeIndexGranularityInfo::getMarksExtensionFromFilesystem(data_part_storage);
|
||||
auto mrk_ext = MergeTreeIndexGranularityInfo::getMarksExtensionFromFilesystem(*data_part_storage);
|
||||
|
||||
if (mrk_ext)
|
||||
{
|
||||
@ -2943,12 +2953,11 @@ MergeTreeData::DataPartsVector MergeTreeData::getActivePartsToReplace(
|
||||
bool MergeTreeData::renameTempPartAndAdd(
|
||||
MutableDataPartPtr & part,
|
||||
Transaction & out_transaction,
|
||||
DataPartStorageBuilderPtr builder,
|
||||
DataPartsLock & lock)
|
||||
{
|
||||
DataPartsVector covered_parts;
|
||||
|
||||
if (!renameTempPartAndReplaceImpl(part, out_transaction, lock, builder, &covered_parts))
|
||||
if (!renameTempPartAndReplaceImpl(part, out_transaction, lock, &covered_parts))
|
||||
return false;
|
||||
|
||||
if (!covered_parts.empty())
|
||||
@ -2982,32 +2991,31 @@ void MergeTreeData::checkPartCanBeAddedToTable(MutableDataPartPtr & part, DataPa
|
||||
}
|
||||
}
|
||||
|
||||
void MergeTreeData::preparePartForCommit(MutableDataPartPtr & part, Transaction & out_transaction, DataPartStorageBuilderPtr builder)
|
||||
void MergeTreeData::preparePartForCommit(MutableDataPartPtr & part, Transaction & out_transaction)
|
||||
{
|
||||
part->is_temp = false;
|
||||
part->setState(DataPartState::PreActive);
|
||||
|
||||
assert([&]()
|
||||
{
|
||||
String dir_name = fs::path(part->data_part_storage->getRelativePath()).filename();
|
||||
String dir_name = fs::path(part->getDataPartStorage().getRelativePath()).filename();
|
||||
bool may_be_cleaned_up = dir_name.starts_with("tmp_") || dir_name.starts_with("tmp-fetch_");
|
||||
return !may_be_cleaned_up || temporary_parts.contains(dir_name);
|
||||
}());
|
||||
|
||||
part->renameTo(part->name, true, builder);
|
||||
part->renameTo(part->name, true);
|
||||
|
||||
data_parts_indexes.insert(part);
|
||||
out_transaction.addPart(part, builder);
|
||||
out_transaction.addPart(part);
|
||||
}
|
||||
|
||||
bool MergeTreeData::renameTempPartAndReplaceImpl(
|
||||
MutableDataPartPtr & part,
|
||||
Transaction & out_transaction,
|
||||
DataPartsLock & lock,
|
||||
DataPartStorageBuilderPtr builder,
|
||||
DataPartsVector * out_covered_parts)
|
||||
{
|
||||
LOG_TRACE(log, "Renaming temporary part {} to {}.", part->data_part_storage->getPartDirectory(), part->name);
|
||||
LOG_TRACE(log, "Renaming temporary part {} to {}.", part->getDataPartStorage().getPartDirectory(), part->name);
|
||||
|
||||
if (&out_transaction.data != this)
|
||||
throw Exception("MergeTreeData::Transaction for one table cannot be used with another. It is a bug.",
|
||||
@ -3029,7 +3037,7 @@ bool MergeTreeData::renameTempPartAndReplaceImpl(
|
||||
|
||||
/// All checks are passed. Now we can rename the part on disk.
|
||||
/// So, we maintain invariant: if a non-temporary part in filesystem then it is in data_parts
|
||||
preparePartForCommit(part, out_transaction, builder);
|
||||
preparePartForCommit(part, out_transaction);
|
||||
|
||||
if (out_covered_parts)
|
||||
{
|
||||
@ -3045,21 +3053,19 @@ bool MergeTreeData::renameTempPartAndReplaceImpl(
|
||||
MergeTreeData::DataPartsVector MergeTreeData::renameTempPartAndReplaceUnlocked(
|
||||
MutableDataPartPtr & part,
|
||||
Transaction & out_transaction,
|
||||
DataPartStorageBuilderPtr builder,
|
||||
DataPartsLock & lock)
|
||||
{
|
||||
DataPartsVector covered_parts;
|
||||
renameTempPartAndReplaceImpl(part, out_transaction, lock, builder, &covered_parts);
|
||||
renameTempPartAndReplaceImpl(part, out_transaction, lock, &covered_parts);
|
||||
return covered_parts;
|
||||
}
|
||||
|
||||
MergeTreeData::DataPartsVector MergeTreeData::renameTempPartAndReplace(
|
||||
MutableDataPartPtr & part,
|
||||
Transaction & out_transaction,
|
||||
DataPartStorageBuilderPtr builder)
|
||||
Transaction & out_transaction)
|
||||
{
|
||||
auto part_lock = lockParts();
|
||||
return renameTempPartAndReplaceUnlocked(part, out_transaction, builder, part_lock);
|
||||
return renameTempPartAndReplaceUnlocked(part, out_transaction, part_lock);
|
||||
}
|
||||
|
||||
void MergeTreeData::removePartsFromWorkingSet(MergeTreeTransaction * txn, const MergeTreeData::DataPartsVector & remove, bool clear_without_timeout, DataPartsLock & acquired_lock)
|
||||
@ -3240,9 +3246,9 @@ void MergeTreeData::outdateBrokenPartAndCloneToDetached(const DataPartPtr & part
|
||||
{
|
||||
auto metadata_snapshot = getInMemoryMetadataPtr();
|
||||
if (prefix.empty())
|
||||
LOG_INFO(log, "Cloning part {} to {} and making it obsolete.", part_to_detach->data_part_storage->getPartDirectory(), part_to_detach->name);
|
||||
LOG_INFO(log, "Cloning part {} to {} and making it obsolete.", part_to_detach->getDataPartStorage().getPartDirectory(), part_to_detach->name);
|
||||
else
|
||||
LOG_INFO(log, "Cloning part {} to {}_{} and making it obsolete.", part_to_detach->data_part_storage->getPartDirectory(), prefix, part_to_detach->name);
|
||||
LOG_INFO(log, "Cloning part {} to {}_{} and making it obsolete.", part_to_detach->getDataPartStorage().getPartDirectory(), prefix, part_to_detach->name);
|
||||
|
||||
part_to_detach->makeCloneInDetached(prefix, metadata_snapshot);
|
||||
|
||||
@ -3254,9 +3260,9 @@ void MergeTreeData::outdateBrokenPartAndCloneToDetached(const DataPartPtr & part
|
||||
void MergeTreeData::forcefullyMovePartToDetachedAndRemoveFromMemory(const MergeTreeData::DataPartPtr & part_to_detach, const String & prefix, bool restore_covered)
|
||||
{
|
||||
if (prefix.empty())
|
||||
LOG_INFO(log, "Renaming {} to {} and forgetting it.", part_to_detach->data_part_storage->getPartDirectory(), part_to_detach->name);
|
||||
LOG_INFO(log, "Renaming {} to {} and forgetting it.", part_to_detach->getDataPartStorage().getPartDirectory(), part_to_detach->name);
|
||||
else
|
||||
LOG_INFO(log, "Renaming {} to {}_{} and forgetting it.", part_to_detach->data_part_storage->getPartDirectory(), prefix, part_to_detach->name);
|
||||
LOG_INFO(log, "Renaming {} to {}_{} and forgetting it.", part_to_detach->getDataPartStorage().getPartDirectory(), prefix, part_to_detach->name);
|
||||
|
||||
auto lock = lockParts();
|
||||
bool removed_active_part = false;
|
||||
@ -3279,11 +3285,7 @@ void MergeTreeData::forcefullyMovePartToDetachedAndRemoveFromMemory(const MergeT
|
||||
}
|
||||
|
||||
modifyPartState(it_part, DataPartState::Deleting);
|
||||
|
||||
auto builder = part->data_part_storage->getBuilder();
|
||||
part->renameToDetached(prefix, builder);
|
||||
builder->commit();
|
||||
|
||||
preparePartForRemoval(part)->renameToDetached(prefix);
|
||||
data_parts_indexes.erase(it_part);
|
||||
|
||||
if (restore_covered && part->info.level == 0)
|
||||
@ -3437,7 +3439,7 @@ void MergeTreeData::tryRemovePartImmediately(DataPartPtr && part)
|
||||
|
||||
try
|
||||
{
|
||||
part_to_delete->remove();
|
||||
preparePartForRemoval(part_to_delete)->remove();
|
||||
}
|
||||
catch (...)
|
||||
{
|
||||
@ -3647,9 +3649,9 @@ void MergeTreeData::swapActivePart(MergeTreeData::DataPartPtr part_copy)
|
||||
/// when allow_remote_fs_zero_copy_replication turned on and off again
|
||||
original_active_part->force_keep_shared_data = false;
|
||||
|
||||
if (original_active_part->data_part_storage->supportZeroCopyReplication() &&
|
||||
part_copy->data_part_storage->supportZeroCopyReplication() &&
|
||||
original_active_part->data_part_storage->getUniqueId() == part_copy->data_part_storage->getUniqueId())
|
||||
if (original_active_part->getDataPartStorage().supportZeroCopyReplication() &&
|
||||
part_copy->getDataPartStorage().supportZeroCopyReplication() &&
|
||||
original_active_part->getDataPartStorage().getUniqueId() == part_copy->getDataPartStorage().getUniqueId())
|
||||
{
|
||||
/// May be when several volumes use the same S3/HDFS storage
|
||||
original_active_part->force_keep_shared_data = true;
|
||||
@ -3669,7 +3671,7 @@ void MergeTreeData::swapActivePart(MergeTreeData::DataPartPtr part_copy)
|
||||
/// All other locks are taken in StorageReplicatedMergeTree
|
||||
lockSharedData(*part_copy);
|
||||
|
||||
original_active_part->data_part_storage->writeDeleteOnDestroyMarker(log);
|
||||
preparePartForRemoval(original_active_part)->writeDeleteOnDestroyMarker();
|
||||
return;
|
||||
}
|
||||
}
|
||||
@ -3803,9 +3805,9 @@ MergeTreeData::DataPartPtr MergeTreeData::getPartIfExists(const String & part_na
|
||||
static void loadPartAndFixMetadataImpl(MergeTreeData::MutableDataPartPtr part)
|
||||
{
|
||||
part->loadColumnsChecksumsIndexes(false, true);
|
||||
part->modification_time = part->data_part_storage->getLastModified().epochTime();
|
||||
part->data_part_storage->removeDeleteOnDestroyMarker();
|
||||
part->data_part_storage->removeVersionMetadata();
|
||||
part->modification_time = part->getDataPartStorage().getLastModified().epochTime();
|
||||
part->removeDeleteOnDestroyMarker();
|
||||
part->removeVersionMetadata();
|
||||
}
|
||||
|
||||
void MergeTreeData::calculateColumnAndSecondaryIndexSizesImpl()
|
||||
@ -3965,7 +3967,7 @@ void MergeTreeData::movePartitionToDisk(const ASTPtr & partition, const String &
|
||||
auto disk = getStoragePolicy()->getDiskByName(name);
|
||||
std::erase_if(parts, [&](auto part_ptr)
|
||||
{
|
||||
return part_ptr->data_part_storage->getDiskName() == disk->getName();
|
||||
return part_ptr->getDataPartStorage().getDiskName() == disk->getName();
|
||||
});
|
||||
|
||||
if (parts.empty())
|
||||
@ -4015,7 +4017,7 @@ void MergeTreeData::movePartitionToVolume(const ASTPtr & partition, const String
|
||||
{
|
||||
for (const auto & disk : volume->getDisks())
|
||||
{
|
||||
if (part_ptr->data_part_storage->getDiskName() == disk->getName())
|
||||
if (part_ptr->getDataPartStorage().getDiskName() == disk->getName())
|
||||
{
|
||||
return true;
|
||||
}
|
||||
@ -4212,7 +4214,7 @@ BackupEntries MergeTreeData::backupParts(const DataPartsVector & data_parts, con
|
||||
make_temporary_hard_links = false;
|
||||
hold_storage_and_part_ptrs = true;
|
||||
}
|
||||
else if (supportsReplication() && part->data_part_storage->supportZeroCopyReplication() && getSettings()->allow_remote_fs_zero_copy_replication)
|
||||
else if (supportsReplication() && part->getDataPartStorage().supportZeroCopyReplication() && getSettings()->allow_remote_fs_zero_copy_replication)
|
||||
{
|
||||
/// Hard links don't work correctly with zero copy replication.
|
||||
make_temporary_hard_links = false;
|
||||
@ -4224,7 +4226,7 @@ BackupEntries MergeTreeData::backupParts(const DataPartsVector & data_parts, con
|
||||
table_lock = lockForShare(local_context->getCurrentQueryId(), local_context->getSettingsRef().lock_acquire_timeout);
|
||||
|
||||
BackupEntries backup_entries_from_part;
|
||||
part->data_part_storage->backup(
|
||||
part->getDataPartStorage().backup(
|
||||
part->checksums,
|
||||
part->getFileNamesWithoutChecksums(),
|
||||
data_path_in_backup,
|
||||
@ -4235,7 +4237,7 @@ BackupEntries MergeTreeData::backupParts(const DataPartsVector & data_parts, con
|
||||
auto projection_parts = part->getProjectionParts();
|
||||
for (const auto & [projection_name, projection_part] : projection_parts)
|
||||
{
|
||||
projection_part->data_part_storage->backup(
|
||||
projection_part->getDataPartStorage().backup(
|
||||
projection_part->checksums,
|
||||
projection_part->getFileNamesWithoutChecksums(),
|
||||
fs::path{data_path_in_backup} / part->name,
|
||||
@ -4911,22 +4913,16 @@ ReservationPtr MergeTreeData::reserveSpace(UInt64 expected_size, SpacePtr space)
|
||||
return checkAndReturnReservation(expected_size, std::move(reservation));
|
||||
}
|
||||
|
||||
ReservationPtr MergeTreeData::reserveSpace(UInt64 expected_size, const DataPartStoragePtr & data_part_storage)
|
||||
ReservationPtr MergeTreeData::reserveSpace(UInt64 expected_size, const IDataPartStorage & data_part_storage)
|
||||
{
|
||||
expected_size = std::max(RESERVATION_MIN_ESTIMATION_SIZE, expected_size);
|
||||
return data_part_storage->reserve(expected_size);
|
||||
return data_part_storage.reserve(expected_size);
|
||||
}
|
||||
|
||||
ReservationPtr MergeTreeData::reserveSpace(UInt64 expected_size, const DataPartStorageBuilderPtr & data_part_storage_builder)
|
||||
ReservationPtr MergeTreeData::tryReserveSpace(UInt64 expected_size, const IDataPartStorage & data_part_storage)
|
||||
{
|
||||
expected_size = std::max(RESERVATION_MIN_ESTIMATION_SIZE, expected_size);
|
||||
return data_part_storage_builder->reserve(expected_size);
|
||||
}
|
||||
|
||||
ReservationPtr MergeTreeData::tryReserveSpace(UInt64 expected_size, const DataPartStoragePtr & data_part_storage)
|
||||
{
|
||||
expected_size = std::max(RESERVATION_MIN_ESTIMATION_SIZE, expected_size);
|
||||
return data_part_storage->tryReserve(expected_size);
|
||||
return data_part_storage.tryReserve(expected_size);
|
||||
}
|
||||
|
||||
ReservationPtr MergeTreeData::tryReserveSpace(UInt64 expected_size, SpacePtr space)
|
||||
@ -5063,7 +5059,7 @@ bool MergeTreeData::shouldPerformTTLMoveOnInsert(const SpacePtr & move_destinati
|
||||
if (move_destination->isDisk())
|
||||
{
|
||||
auto disk = std::static_pointer_cast<IDisk>(move_destination);
|
||||
if (auto volume = getStoragePolicy()->tryGetVolumeByDisk(disk))
|
||||
if (auto volume = getStoragePolicy()->tryGetVolumeByDiskName(disk->getName()))
|
||||
return volume->perform_ttl_move_on_insert;
|
||||
}
|
||||
return false;
|
||||
@ -5075,11 +5071,11 @@ bool MergeTreeData::isPartInTTLDestination(const TTLDescription & ttl, const IMe
|
||||
if (ttl.destination_type == DataDestinationType::VOLUME)
|
||||
{
|
||||
for (const auto & disk : policy->getVolumeByName(ttl.destination_name)->getDisks())
|
||||
if (disk->getName() == part.data_part_storage->getDiskName())
|
||||
if (disk->getName() == part.getDataPartStorage().getDiskName())
|
||||
return true;
|
||||
}
|
||||
else if (ttl.destination_type == DataDestinationType::DISK)
|
||||
return policy->getDiskByName(ttl.destination_name)->getName() == part.data_part_storage->getDiskName();
|
||||
return policy->getDiskByName(ttl.destination_name)->getName() == part.getDataPartStorage().getDiskName();
|
||||
return false;
|
||||
}
|
||||
|
||||
@ -5151,7 +5147,7 @@ void MergeTreeData::Transaction::rollbackPartsToTemporaryState()
|
||||
WriteBufferFromOwnString buf;
|
||||
buf << " Rollbacking parts state to temporary and removing from working set:";
|
||||
for (const auto & part : precommitted_parts)
|
||||
buf << " " << part->data_part_storage->getPartDirectory();
|
||||
buf << " " << part->getDataPartStorage().getPartDirectory();
|
||||
buf << ".";
|
||||
LOG_DEBUG(data.log, "Undoing transaction.{}", buf.str());
|
||||
|
||||
@ -5162,12 +5158,11 @@ void MergeTreeData::Transaction::rollbackPartsToTemporaryState()
|
||||
clear();
|
||||
}
|
||||
|
||||
void MergeTreeData::Transaction::addPart(MutableDataPartPtr & part, DataPartStorageBuilderPtr builder)
|
||||
void MergeTreeData::Transaction::addPart(MutableDataPartPtr & part)
|
||||
{
|
||||
precommitted_parts.insert(part);
|
||||
if (asInMemoryPart(part))
|
||||
has_in_memory_parts = true;
|
||||
part_builders.push_back(builder);
|
||||
}
|
||||
|
||||
void MergeTreeData::Transaction::rollback()
|
||||
@ -5177,7 +5172,7 @@ void MergeTreeData::Transaction::rollback()
|
||||
WriteBufferFromOwnString buf;
|
||||
buf << " Removing parts:";
|
||||
for (const auto & part : precommitted_parts)
|
||||
buf << " " << part->data_part_storage->getPartDirectory();
|
||||
buf << " " << part->getDataPartStorage().getPartDirectory();
|
||||
buf << ".";
|
||||
LOG_DEBUG(data.log, "Undoing transaction.{}", buf.str());
|
||||
|
||||
@ -5205,8 +5200,9 @@ MergeTreeData::DataPartsVector MergeTreeData::Transaction::commit(MergeTreeData:
|
||||
auto parts_lock = acquired_parts_lock ? MergeTreeData::DataPartsLock() : data.lockParts();
|
||||
auto * owing_parts_lock = acquired_parts_lock ? acquired_parts_lock : &parts_lock;
|
||||
|
||||
for (auto & builder : part_builders)
|
||||
builder->commit();
|
||||
for (const auto & part : precommitted_parts)
|
||||
if (part->getDataPartStorage().hasActiveTransaction())
|
||||
part->getDataPartStorage().commitTransaction();
|
||||
|
||||
bool commit_to_wal = has_in_memory_parts && settings->in_memory_parts_enable_wal;
|
||||
if (txn || commit_to_wal)
|
||||
@ -5215,7 +5211,7 @@ MergeTreeData::DataPartsVector MergeTreeData::Transaction::commit(MergeTreeData:
|
||||
if (commit_to_wal)
|
||||
wal = data.getWriteAheadLog();
|
||||
|
||||
for (const DataPartPtr & part : precommitted_parts)
|
||||
for (const auto & part : precommitted_parts)
|
||||
{
|
||||
if (txn)
|
||||
{
|
||||
@ -5240,7 +5236,7 @@ MergeTreeData::DataPartsVector MergeTreeData::Transaction::commit(MergeTreeData:
|
||||
size_t reduce_rows = 0;
|
||||
size_t reduce_parts = 0;
|
||||
|
||||
for (const DataPartPtr & part : precommitted_parts)
|
||||
for (const auto & part : precommitted_parts)
|
||||
{
|
||||
DataPartPtr covering_part;
|
||||
DataPartsVector covered_parts = data.getActivePartsToReplace(part->info, part->name, covering_part, *owing_parts_lock);
|
||||
@ -6232,7 +6228,7 @@ std::pair<MergeTreeData::MutableDataPartPtr, scope_guard> MergeTreeData::cloneAn
|
||||
bool does_storage_policy_allow_same_disk = false;
|
||||
for (const DiskPtr & disk : getStoragePolicy()->getDisks())
|
||||
{
|
||||
if (disk->getName() == src_part->data_part_storage->getDiskName())
|
||||
if (disk->getName() == src_part->getDataPartStorage().getDiskName())
|
||||
{
|
||||
does_storage_policy_allow_same_disk = true;
|
||||
break;
|
||||
@ -6242,7 +6238,7 @@ std::pair<MergeTreeData::MutableDataPartPtr, scope_guard> MergeTreeData::cloneAn
|
||||
throw Exception(
|
||||
ErrorCodes::BAD_ARGUMENTS,
|
||||
"Could not clone and load part {} because disk does not belong to storage policy",
|
||||
quoteString(src_part->data_part_storage->getFullPath()));
|
||||
quoteString(src_part->getDataPartStorage().getFullPath()));
|
||||
|
||||
String dst_part_name = src_part->getNewName(dst_part_info);
|
||||
assert(!tmp_part_prefix.empty());
|
||||
@ -6250,9 +6246,8 @@ std::pair<MergeTreeData::MutableDataPartPtr, scope_guard> MergeTreeData::cloneAn
|
||||
auto temporary_directory_lock = getTemporaryPartDirectoryHolder(tmp_dst_part_name);
|
||||
|
||||
/// Why it is needed if we only hardlink files?
|
||||
auto reservation = src_part->data_part_storage->reserve(src_part->getBytesOnDisk());
|
||||
|
||||
auto src_part_storage = src_part->data_part_storage;
|
||||
auto reservation = src_part->getDataPartStorage().reserve(src_part->getBytesOnDisk());
|
||||
auto src_part_storage = src_part->getDataPartStoragePtr();
|
||||
|
||||
/// If source part is in memory, flush it to disk and clone it already in on-disk format
|
||||
if (auto src_part_in_memory = asInMemoryPart(src_part))
|
||||
@ -6279,7 +6274,7 @@ std::pair<MergeTreeData::MutableDataPartPtr, scope_guard> MergeTreeData::cloneAn
|
||||
hardlinked_files->source_part_name = src_part->name;
|
||||
hardlinked_files->source_table_shared_id = src_part->storage.getTableSharedID();
|
||||
|
||||
for (auto it = src_part->data_part_storage->iterate(); it->isValid(); it->next())
|
||||
for (auto it = src_part->getDataPartStorage().iterate(); it->isValid(); it->next())
|
||||
{
|
||||
if (!files_to_copy_instead_of_hardlinks.contains(it->name())
|
||||
&& it->name() != IMergeTreeDataPart::DELETE_ON_DESTROY_MARKER_FILE_NAME
|
||||
@ -6338,14 +6333,14 @@ Strings MergeTreeData::getDataPaths() const
|
||||
|
||||
void MergeTreeData::reportBrokenPart(MergeTreeData::DataPartPtr & data_part) const
|
||||
{
|
||||
if (data_part->data_part_storage && data_part->data_part_storage->isBroken())
|
||||
if (data_part->getDataPartStorage().isBroken())
|
||||
{
|
||||
auto parts = getDataPartsForInternalUsage();
|
||||
LOG_WARNING(log, "Scanning parts to recover on broken disk {}@{}.", data_part->data_part_storage->getDiskName(), data_part->data_part_storage->getDiskPath());
|
||||
LOG_WARNING(log, "Scanning parts to recover on broken disk {}@{}.", data_part->getDataPartStorage().getDiskName(), data_part->getDataPartStorage().getDiskPath());
|
||||
|
||||
for (const auto & part : parts)
|
||||
{
|
||||
if (part->data_part_storage && part->data_part_storage->getDiskName() == data_part->data_part_storage->getDiskName())
|
||||
if (part->getDataPartStorage().getDiskName() == data_part->getDataPartStorage().getDiskName())
|
||||
broken_part_callback(part->name);
|
||||
}
|
||||
}
|
||||
@ -6436,7 +6431,7 @@ PartitionCommandsResultInfo MergeTreeData::freezePartitionsByMatcher(
|
||||
|
||||
LOG_DEBUG(log, "Freezing part {} snapshot will be placed at {}", part->name, backup_path);
|
||||
|
||||
auto data_part_storage = part->data_part_storage;
|
||||
auto data_part_storage = part->getDataPartStoragePtr();
|
||||
String src_part_path = data_part_storage->getRelativePath();
|
||||
String backup_part_path = fs::path(backup_path) / relative_data_path;
|
||||
if (auto part_in_memory = asInMemoryPart(part))
|
||||
@ -6450,12 +6445,12 @@ PartitionCommandsResultInfo MergeTreeData::freezePartitionsByMatcher(
|
||||
|
||||
// Store metadata for replicated table.
|
||||
// Do nothing for non-replicated.
|
||||
createAndStoreFreezeMetadata(disk, part, fs::path(backup_part_path) / part->data_part_storage->getPartDirectory());
|
||||
createAndStoreFreezeMetadata(disk, part, fs::path(backup_part_path) / part->getDataPartStorage().getPartDirectory());
|
||||
};
|
||||
|
||||
auto new_storage = data_part_storage->freeze(
|
||||
backup_part_path,
|
||||
part->data_part_storage->getPartDirectory(),
|
||||
part->getDataPartStorage().getPartDirectory(),
|
||||
/*make_source_readonly*/ true,
|
||||
callback,
|
||||
/*copy_instead_of_hardlink*/ false,
|
||||
@ -6577,8 +6572,8 @@ try
|
||||
|
||||
if (result_part)
|
||||
{
|
||||
part_log_elem.disk_name = result_part->data_part_storage->getDiskName();
|
||||
part_log_elem.path_on_disk = result_part->data_part_storage->getFullPath();
|
||||
part_log_elem.disk_name = result_part->getDataPartStorage().getDiskName();
|
||||
part_log_elem.path_on_disk = result_part->getDataPartStorage().getFullPath();
|
||||
part_log_elem.bytes_compressed_on_disk = result_part->getBytesOnDisk();
|
||||
part_log_elem.rows = result_part->rows_count;
|
||||
part_log_elem.part_type = result_part->getType();
|
||||
@ -6734,7 +6729,7 @@ bool MergeTreeData::moveParts(const CurrentlyMovingPartsTaggerPtr & moving_tagge
|
||||
for (const auto & moving_part : moving_tagger->parts_to_move)
|
||||
{
|
||||
Stopwatch stopwatch;
|
||||
DataPartPtr cloned_part;
|
||||
MutableDataPartPtr cloned_part;
|
||||
|
||||
auto write_part_log = [&](const ExecutionStatus & execution_status)
|
||||
{
|
||||
@ -6997,7 +6992,7 @@ ReservationPtr MergeTreeData::balancedReservation(
|
||||
if (part->isStoredOnDisk() && part->getBytesOnDisk() >= min_bytes_to_rebalance_partition_over_jbod
|
||||
&& part_info.partition_id == part->info.partition_id)
|
||||
{
|
||||
auto name = part->data_part_storage->getDiskName();
|
||||
auto name = part->getDataPartStorage().getDiskName();
|
||||
auto it = disk_occupation.find(name);
|
||||
if (it != disk_occupation.end())
|
||||
{
|
||||
|
||||
@ -214,6 +214,7 @@ public:
|
||||
};
|
||||
|
||||
using DataParts = std::set<DataPartPtr, LessDataPart>;
|
||||
using MutableDataParts = std::set<MutableDataPartPtr, LessDataPart>;
|
||||
using DataPartsVector = std::vector<DataPartPtr>;
|
||||
|
||||
using DataPartsLock = std::unique_lock<std::mutex>;
|
||||
@ -225,15 +226,15 @@ public:
|
||||
/// After this method setColumns must be called
|
||||
MutableDataPartPtr createPart(const String & name,
|
||||
MergeTreeDataPartType type, const MergeTreePartInfo & part_info,
|
||||
const DataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part = nullptr) const;
|
||||
const MutableDataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part = nullptr) const;
|
||||
|
||||
/// Create part, that already exists on filesystem.
|
||||
/// After this methods 'loadColumnsChecksumsIndexes' must be called.
|
||||
MutableDataPartPtr createPart(const String & name,
|
||||
const DataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part = nullptr) const;
|
||||
const MutableDataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part = nullptr) const;
|
||||
|
||||
MutableDataPartPtr createPart(const String & name, const MergeTreePartInfo & part_info,
|
||||
const DataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part = nullptr) const;
|
||||
const MutableDataPartStoragePtr & data_part_storage, const IMergeTreeDataPart * parent_part = nullptr) const;
|
||||
|
||||
/// Auxiliary object to add a set of parts into the working set in two steps:
|
||||
/// * First, as PreActive parts (the parts are ready, but not yet in the active set).
|
||||
@ -247,7 +248,7 @@ public:
|
||||
|
||||
DataPartsVector commit(MergeTreeData::DataPartsLock * acquired_parts_lock = nullptr);
|
||||
|
||||
void addPart(MutableDataPartPtr & part, DataPartStorageBuilderPtr builder);
|
||||
void addPart(MutableDataPartPtr & part);
|
||||
|
||||
void rollback();
|
||||
|
||||
@ -275,9 +276,8 @@ public:
|
||||
|
||||
MergeTreeData & data;
|
||||
MergeTreeTransaction * txn;
|
||||
DataParts precommitted_parts;
|
||||
std::vector<DataPartStorageBuilderPtr> part_builders;
|
||||
DataParts locked_parts;
|
||||
MutableDataParts precommitted_parts;
|
||||
MutableDataParts locked_parts;
|
||||
bool has_in_memory_parts = false;
|
||||
|
||||
void clear();
|
||||
@ -414,9 +414,8 @@ public:
|
||||
SelectQueryInfo & info) const override;
|
||||
|
||||
ReservationPtr reserveSpace(UInt64 expected_size, VolumePtr & volume) const;
|
||||
static ReservationPtr tryReserveSpace(UInt64 expected_size, const DataPartStoragePtr & data_part_storage);
|
||||
static ReservationPtr reserveSpace(UInt64 expected_size, const DataPartStoragePtr & data_part_storage);
|
||||
static ReservationPtr reserveSpace(UInt64 expected_size, const DataPartStorageBuilderPtr & data_part_storage_builder);
|
||||
static ReservationPtr tryReserveSpace(UInt64 expected_size, const IDataPartStorage & data_part_storage);
|
||||
static ReservationPtr reserveSpace(UInt64 expected_size, const IDataPartStorage & data_part_storage);
|
||||
|
||||
static bool partsContainSameProjections(const DataPartPtr & left, const DataPartPtr & right);
|
||||
|
||||
@ -555,21 +554,18 @@ public:
|
||||
bool renameTempPartAndAdd(
|
||||
MutableDataPartPtr & part,
|
||||
Transaction & transaction,
|
||||
DataPartStorageBuilderPtr builder,
|
||||
DataPartsLock & lock);
|
||||
|
||||
/// The same as renameTempPartAndAdd but the block range of the part can contain existing parts.
|
||||
/// Returns all parts covered by the added part (in ascending order).
|
||||
DataPartsVector renameTempPartAndReplace(
|
||||
MutableDataPartPtr & part,
|
||||
Transaction & out_transaction,
|
||||
DataPartStorageBuilderPtr builder);
|
||||
Transaction & out_transaction);
|
||||
|
||||
/// Unlocked version of previous one. Useful when added multiple parts with a single lock.
|
||||
DataPartsVector renameTempPartAndReplaceUnlocked(
|
||||
MutableDataPartPtr & part,
|
||||
Transaction & out_transaction,
|
||||
DataPartStorageBuilderPtr builder,
|
||||
DataPartsLock & lock);
|
||||
|
||||
/// Remove parts from working set immediately (without wait for background
|
||||
@ -979,7 +975,7 @@ public:
|
||||
|
||||
/// Fetch part only if some replica has it on shared storage like S3
|
||||
/// Overridden in StorageReplicatedMergeTree
|
||||
virtual DataPartStoragePtr tryToFetchIfShared(const IMergeTreeDataPart &, const DiskPtr &, const String &) { return nullptr; }
|
||||
virtual MutableDataPartStoragePtr tryToFetchIfShared(const IMergeTreeDataPart &, const DiskPtr &, const String &) { return nullptr; }
|
||||
|
||||
/// Check shared data usage on other replicas for detached/freezed part
|
||||
/// Remove local files and remote files if needed
|
||||
@ -1264,13 +1260,12 @@ protected:
|
||||
static void incrementMergedPartsProfileEvent(MergeTreeDataPartType type);
|
||||
|
||||
private:
|
||||
|
||||
/// Checking that candidate part doesn't break invariants: correct partition and doesn't exist already
|
||||
void checkPartCanBeAddedToTable(MutableDataPartPtr & part, DataPartsLock & lock) const;
|
||||
|
||||
/// Preparing itself to be committed in memory: fill some fields inside part, add it to data_parts_indexes
|
||||
/// in precommitted state and to transaction
|
||||
void preparePartForCommit(MutableDataPartPtr & part, Transaction & out_transaction, DataPartStorageBuilderPtr builder);
|
||||
void preparePartForCommit(MutableDataPartPtr & part, Transaction & out_transaction);
|
||||
|
||||
/// Low-level method for preparing parts for commit (in-memory).
|
||||
/// FIXME Merge MergeTreeTransaction and Transaction
|
||||
@ -1278,7 +1273,6 @@ private:
|
||||
MutableDataPartPtr & part,
|
||||
Transaction & out_transaction,
|
||||
DataPartsLock & lock,
|
||||
DataPartStorageBuilderPtr builder,
|
||||
DataPartsVector * out_covered_parts);
|
||||
|
||||
/// RAII Wrapper for atomic work with currently moving parts
|
||||
@ -1334,8 +1328,8 @@ private:
|
||||
virtual std::unique_ptr<MergeTreeSettings> getDefaultSettings() const = 0;
|
||||
|
||||
void loadDataPartsFromDisk(
|
||||
DataPartsVector & broken_parts_to_detach,
|
||||
DataPartsVector & duplicate_parts_to_remove,
|
||||
MutableDataPartsVector & broken_parts_to_detach,
|
||||
MutableDataPartsVector & duplicate_parts_to_remove,
|
||||
ThreadPool & pool,
|
||||
size_t num_parts,
|
||||
std::queue<std::vector<std::pair<String, DiskPtr>>> & parts_queue,
|
||||
@ -1343,8 +1337,7 @@ private:
|
||||
const MergeTreeSettingsPtr & settings);
|
||||
|
||||
void loadDataPartsFromWAL(
|
||||
DataPartsVector & broken_parts_to_detach,
|
||||
DataPartsVector & duplicate_parts_to_remove,
|
||||
MutableDataPartsVector & duplicate_parts_to_remove,
|
||||
MutableDataPartsVector & parts_from_wal);
|
||||
|
||||
/// Create zero-copy exclusive lock for part and disk. Useful for coordination of
|
||||
@ -1356,6 +1349,8 @@ private:
|
||||
/// Otherwise, in non-parallel case will break and return.
|
||||
void clearPartsFromFilesystemImpl(const DataPartsVector & parts, NameSet * part_names_succeed);
|
||||
|
||||
static MutableDataPartPtr preparePartForRemoval(const DataPartPtr & part);
|
||||
|
||||
TemporaryParts temporary_parts;
|
||||
};
|
||||
|
||||
|
||||
Some files were not shown because too many files have changed in this diff Show More
Loading…
Reference in New Issue
Block a user