Commit Graph

12436 Commits

Author SHA1 Message Date
Alexey Zatelepin
6bb34a4494 Merge branch 'master' into arbitrary-partitioning
Conflicts:
	dbms/src/Storages/StorageFactory.cpp
2017-10-24 17:23:18 +03:00
Alexey Zatelepin
b6b5d1cf25 store rows count for part on disk to allow tables without fixed size columns [#CLICKHOUSE-3000] 2017-10-24 17:11:53 +03:00
alexey-milovidov
d5cf664297 Update build.rst 2017-10-24 02:50:54 +03:00
proller
82134e44cb Split ComplexKeyCacheDictionary to faster compile (part2) 2017-10-24 00:26:42 +03:00
proller
86a587e482 Fix test (supress stderr) 2017-10-23 21:40:07 +03:00
alexey-milovidov
3c2155df0b Update StorageKafka.cpp 2017-10-23 20:47:33 +03:00
alexey-milovidov
139e2080ae Update StorageKafka.cpp 2017-10-23 20:47:33 +03:00
alexey-milovidov
5bfb4c213c Update JSONEachRowRowInputStream.cpp 2017-10-23 20:47:33 +03:00
Marek Vavruša
3926851343 StorageKafka: use config string instead of just server name 2017-10-23 20:47:33 +03:00
Marek Vavruša
a494491f47 contrib: updated librdkafka to v0.11.1@3401fa1
This is the last version with CMake fixes from @proller
2017-10-23 20:47:33 +03:00
Marek Vavruša
a7c391fbf7 StorageKafka: use std::vector instead of stack array for error string buffers 2017-10-23 20:47:33 +03:00
Marek Vavruša
f4c1149606 Settings: renamed schema to format_schema 2017-10-23 20:47:33 +03:00
Marek Vavruša
6be0ffd943 DataStreams/JSONEachRow: read exact number of rows
When using a LIMIT to reduce the number of rows
read, JSONEachRow format would always read an
extra row as it triggered a scan to seek the
EOF or beginning of the next row. This is not
an issue with physical tables, but it is
for streaming tables as an extra row is
unintentionally consumed on each read.
2017-10-23 20:47:33 +03:00
Alexey Milovidov
365f1c2a3b Merge branch 'master' of github.com:yandex/ClickHouse 2017-10-23 20:39:07 +03:00
Alexey Milovidov
c57a62611b Removed StorageTrivialBuffer because it is underdeveloped [#CLICKHOUSE-2]. 2017-10-23 20:38:56 +03:00
proller
13e5e9fbde Fix build in ubuntu xenial in pbuilder 2017-10-23 20:35:43 +03:00
proller
c66228a746 Log query id in executeQuery; Better type mismatch error; change format in report tool (#1372)
* Log query id in executeQuery; Better type mismatch error; change format in report tool

* Better log query_id

* fix message

* Use c++11 thread_local instaed of gcc's __thread

* lock mutex before notifying waiting thread in sync insertion into distributed [#CLICKHOUSE-3379]

* Cmake: fix build without downloaded submodules (#1379)

* fix
2017-10-23 20:32:39 +03:00
proller
ddb0dcd2fd Increase default maximum size of brtie dictionary to 128m ips (#1364)
* Increase default maximum size of brtie dictionary to 128m ips
pbuilder: use newer cmake on ubuntu trusty

* no cmake
2017-10-23 20:19:15 +03:00
Alexey Zatelepin
8ccebefae6 add tests for extended storage definition syntax [#CLICKHOUSE-3000] 2017-10-23 18:21:44 +03:00
alexey-milovidov
6555d08782 Merge pull request #1390 from proller/fix6
Fix: An Error SQL will make server down (#1388)
2017-10-23 18:10:42 +03:00
Alexey Milovidov
d77686685a Using PODArray instead of std::vector in ZooKeeper library [#CLICKHOUSE-2]. 2017-10-23 18:08:31 +03:00
alexey-milovidov
91aa0713e5 Merge pull request #1371 from vavrusa/fix-zookeeper-stack-smashing
ZooKeeper: fixed stack smashing with tryGet()
2017-10-23 17:40:12 +03:00
alexey-milovidov
4f86ec9d58 Update ZooKeeper.cpp 2017-10-23 17:39:29 +03:00
proller
d8038cf746 Fix: An Error SQL will make server down (#1388) 2017-10-23 17:37:46 +03:00
Alex Zatelepin
dd451946cf Update CHANGELOG_RU.md 2017-10-23 17:22:22 +03:00
Nikolai Kochetov
adae8205a7 increased timeout for test_insertion_sync_fails_with_timeout 2017-10-23 16:56:42 +03:00
Nikolai Kochetov
5966574405 fix timeout lock in DistributedBlockOutputStream 2017-10-23 16:47:00 +03:00
Nikolai Kochetov
3317251e2f fix jobs flags in DistributedBlockOutputStream 2017-10-23 15:00:56 +03:00
alexey-milovidov
a7189820a1 Merge pull request #1375 from yandex/fix-hangup-in-sync-writing
lock mutex before notifying waiting thread in sync insertion
2017-10-22 12:45:54 +03:00
alexey-milovidov
c5b5f2b4f8 Merge pull request #1374 from proller/fix4
Use c++11 thread_local instead of gcc's __thread
2017-10-22 12:43:11 +03:00
Alexey Zatelepin
dd42d53856 additional help messages for extended syntax [#CLICKHOUSE-3000] 2017-10-20 23:02:55 +03:00
Alexey Zatelepin
c61d4106e8 allow loading all MergeTree settings from a query [#CLICKHOUSE-3000] 2017-10-20 21:28:45 +03:00
Alexey Zatelepin
a7c85e0c89 save index_granularity to ATTACH query if it wasn't set explicitly [#CLICKHOUSE-3000] 2017-10-20 21:28:44 +03:00
Alexey Zatelepin
1f76900500 move CREATE query rewriting for views to Interpreter [#CLICKHOUSE-2] 2017-10-20 21:28:43 +03:00
Alexey Zatelepin
75c65c7b59 use extended storage definition syntax for MergeTree engines [#CLICKHOUSE-3000] 2017-10-20 21:28:42 +03:00
Alexey Zatelepin
6b9ead4bd8 parse PARTITION BY, ORDER BY, SAMPLE BY, SETTINGS clauses in a storage definition [#CLICKHOUSE-3000] 2017-10-20 21:28:42 +03:00
Alexey Zatelepin
48dea685d0 use more typesafe AST objects in ASTCreateQuery [#CLICKHOUSE-3000] 2017-10-20 21:28:41 +03:00
proller
8dd5c9dac2 Cmake: fix build without downloaded submodules (#1379) 2017-10-20 20:59:44 +03:00
Vitaliy Lyudvichenko
622bd34505 Update CHANGELOG_RU.md 2017-10-20 14:03:01 +03:00
Nikolai Kochetov
4262c00122 lock mutex before notifying waiting thread in sync insertion into distributed [#CLICKHOUSE-3379] 2017-10-19 22:52:45 +03:00
proller
443088ce31 Use c++11 thread_local instaed of gcc's __thread 2017-10-19 22:36:37 +03:00
Alexey Zatelepin
e51f95872f fix backward compatibility for new part format: [#CLICKHOUSE-3000]
* do not write out partition and minmax index files for old MergeTreeData format
* download the part from another replica if there are missing or extra files
2017-10-19 21:24:30 +03:00
Vitaliy Lyudvichenko
24a865a4ff Update ru changelog for 1.1.54301 version. [#CLICKHOUSE-3] 2017-10-19 17:16:43 +03:00
Marek Vavruša
e2a8526990 ZooKeeper: fixed stack smashing with tryGet()
The tryGet() operation creates a 1MB buffer on stack. This may or
may not work depending on the default stack size for threads,
whether the stack protector is enabled or not, recursion depth,
and the actual value size.

This is probably going to slow down some ZK operations, but I don't
see how else this could work reliably with the existing API.
2017-10-18 15:13:42 -07:00
proller
0eb833c433 Fix tcmalloc compile with old clang 2017-10-18 15:24:56 +03:00
alexey-milovidov
3119afbee8 Merge pull request #1366 from avasiliev/master
Added explicit copy constructor for mysqlxx::Null
2017-10-18 12:35:01 +03:00
Alexey Vasiliev
7ec4530354 Added explicit copy constructor for mysqlxx::Null 2017-10-17 20:01:01 +03:00
alexey-milovidov
9183c3f897 Merge pull request #1363 from avasiliev/master
Updated copy/move operations for some types, https://nda.ya.ru/3TWvPE
2017-10-17 19:04:17 +03:00
Alexey Vasiliev
59309432f9 Updated copy/move operations for some types, https://nda.ya.ru/3TWvPE 2017-10-17 14:58:08 +03:00
robot-metrika-test
8a2b7cf0b6 Auto version update to [54300] 2017-10-16 13:49:50 +03:00