Commit Graph

4079 Commits

Author SHA1 Message Date
alesapin
cbae29d26a
Merge pull request #57699 from hanfei1991/hanfei/keeper_soft_limit_ratio
support keeper memory soft limit ratio
2023-12-11 21:43:24 +01:00
Alexey Milovidov
e443567c0e
Merge pull request #52717 from azat/wait-threadpool
[RFC] Correctly wait background threads
2023-12-11 01:06:23 +01:00
Azat Khuzhin
8f5a60eb5a More descriptive comment for destroying context prio global thread pool
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2023-12-10 15:39:43 +01:00
Azat Khuzhin
fcdead023c Proper server termination in case of intermediate failures during initialization
For instance on CI it fails because KeeperDispatch::initialize() failed,
but there is already proper KeeperDispatch::shutdown() in dtor, the
problem is that you first need to shutdown global context, that will
terminate all thread pools, and only after you can join global thread
pool.

CI: https://s3.amazonaws.com/clickhouse-test-reports/52717/6ccbc2ea75c39234dff748cfb147bd44d65d3e75/integration_tests__asan__[3_4].html
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2023-12-10 08:01:40 +01:00
Azat Khuzhin
4f9d90761b Attach tables lazily only for clickhouse-local
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2023-12-09 23:08:29 +01:00
Azat Khuzhin
cc9afe6f6e Replace --no-system-tables with loading virtual tables of system database lazily
First of all the problem is that that simple 'SELECT 1' cannot be runned
without system.one, which makes --no-system-tables almost useless:

    $ ./clickhouse-debug local --no-system-tables -q "select 1"
    Code: 81. DB::Exception: Database system does not exist. (UNKNOWN_DATABASE)

Secondly there are just too much flags, and this one
(--no-system-tables) is too damn specific.

This patch should improve startup time of the clickhouse-local almost
3x in debug builds.

Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2023-12-09 23:08:29 +01:00
Azat Khuzhin
b7edde3621 Shutdown embedded keeper even if it was not started successfully
Since even when it does not, there are still few threads left that
should be joined (request, response, snapshot, snapshot s3):

There are some problems with keeper and invalid config, that got stuck in:

    libc.so.6`___lldb_unnamed_symbol3457 + 231
    libc.so.6`pthread_cond_timedwait + 571
    clickhouse`std::__1::__libcpp_condvar_timedwait[abi:v15000](__cv=0x00007f475c41a708, __m=0x00007f475c41a6b0, __ts=0x00007f45a99f1440) at __threading_support:341:10
    clickhouse`std::__1::condition_variable::__do_timed_wait(this=0x00007f475c41a708, lk=0x00007f45a99f1738, tp=time_point<std::__1::chrono::system_clock, std::__1::chrono::duration<long long, std::__1::ratio<1L, 1000000000L> > > @ 0x00007f45a99f1488) at condition_variable.cpp:76:14
    clickhouse`std::__1::cv_status std::__1::condition_variable::wait_for<long long, std::__1::ratio<1l, 1000000000l> >(this=0x00007f475c41a708, __lk=0x00007f45a99f1738, __d=0x00007f45a99f1590) at __mutex_base:457:9
    clickhouse`void std::__1::condition_variable::__do_timed_wait<std::__1::chrono::steady_clock>(this=0x00007f475c41a708, __lk=0x00007f45a99f1738, __tp=time_point<std::__1::chrono::steady_clock, std::__1::chrono::duration<long long, std::__1::ratio<1L, 1000000000L> > > @ 0x00007f45a99f15a8) at __mutex_base:512:5
    clickhouse`std::__1::cv_status std::__1::condition_variable::wait_until<std::__1::chrono::steady_clock, std::__1::chrono::duration<long long, std::__1::ratio<1l, 1000000000l> > >(this=0x00007f475c41a708, __lk=0x00007f45a99f1738, __t=0x00007f45a99f1698) at __mutex_base:415:5
    clickhouse`bool std::__1::condition_variable::wait_until<std::__1::chrono::steady_clock, std::__1::chrono::duration<long long, std::__1::ratio<1l, 1000000000l> >, bool ConcurrentBoundedQueue<DB::KeeperStorage::RequestForSession>::popImpl<true>(DB::KeeperStorage::RequestForSession&, std::__1::optional<unsigned long>)::'lambda'()>(this=0x00007f475c41a708, __lk=0x00007f45a99f1738, __t=0x00007f45a99f1698, __pred=(unnamed class) @ 0x00007f45a99f1650)::'lambda'()) at __mutex_base:427:13
    clickhouse`bool std::__1::condition_variable::wait_for[abi:v15000]<long long, std::__1::ratio<1l, 1000l>, bool ConcurrentBoundedQueue<DB::KeeperStorage::RequestForSession>::popImpl<true>(DB::KeeperStorage::RequestForSession&, std::__1::optional<unsigned long>)::'lambda'()>(this=0x00007f475c41a708, __lk=0x00007f45a99f1738, __d=0x00007f45a99f1720, __pred=(unnamed class) @ 0x00007f45a99f16b8)::'lambda'()) at __mutex_base:471:12
    clickhouse`bool ConcurrentBoundedQueue<DB::KeeperStorage::RequestForSession>::popImpl<true>(this=0x00007f475c41a680, x=0x00007f45a99f1ba0, timeout_milliseconds= Has Value=true ) at ConcurrentBoundedQueue.h:75:50
    clickhouse`ConcurrentBoundedQueue<DB::KeeperStorage::RequestForSession>::tryPop(this=0x00007f475c41a680, x=0x00007f45a99f1ba0, milliseconds=5000) at ConcurrentBoundedQueue.h:161:16
    clickhouse`DB::KeeperDispatcher::requestThread(this=0x00007f475c8fac18) at KeeperDispatcher.cpp:88:33
    clickhouse`DB::KeeperDispatcher::initialize(this=0x00007f45a99f2020)::$_0::operator()() const at KeeperDispatcher.cpp:349:52
    clickhouse`decltype(__f=0x00007f45a99f2020)::$_0&>()()) std::__1::__invoke[abi:v15000]<DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&) at invoke.h:394:23
    clickhouse`decltype(__f=0x00007f45a99f2020, __t=size=0, (null)=__tuple_indices<> @ 0x00007f45a99f1d28) std::__1::__apply_tuple_impl[abi:v15000]<DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&, std::__1::tuple<>&>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&, std::__1::tuple<>&, std::__1::__tuple_indices<>) at tuple:1789:1
    clickhouse`decltype(__f=0x00007f45a99f2020, __t=size=0) std::__1::apply[abi:v15000]<DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&, std::__1::tuple<>&>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&, std::__1::tuple<>&) at tuple:1798:1
    clickhouse`ThreadFromGlobalPoolImpl<true>::ThreadFromGlobalPoolImpl<DB::KeeperDispatcher::initialize(this=0x00007f475c9703c0)::$_0>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&&)::'lambda'()::operator()() at ThreadPool.h:245:13
    clickhouse`decltype(__f=0x00007f475c9703c0)::$_0>()()) std::__1::__invoke[abi:v15000]<ThreadFromGlobalPoolImpl<true>::ThreadFromGlobalPoolImpl<DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&&)::'lambda'()&>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&&) at invoke.h:394:23
    clickhouse`void std::__1::__invoke_void_return_wrapper<void, true>::__call<ThreadFromGlobalPoolImpl<true>::ThreadFromGlobalPoolImpl<DB::KeeperDispatcher::initialize(__args=0x00007f475c9703c0)::$_0>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&&)::'lambda'()&>(ThreadFromGlobalPoolImpl<true>::ThreadFromGlobalPoolImpl<DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&&)::'lambda'()&) at invoke.h:479:9
    clickhouse`std::__1::__function::__default_alloc_func<ThreadFromGlobalPoolImpl<true>::ThreadFromGlobalPoolImpl<DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&&)::'lambda'(), void ()>::operator(this=0x00007f475c9703c0)[abi:v15000]() at function.h:235:12
    clickhouse`void std::__1::__function::__policy_invoker<void ()>::__call_impl<std::__1::__function::__default_alloc_func<ThreadFromGlobalPoolImpl<true>::ThreadFromGlobalPoolImpl<DB::KeeperDispatcher::initialize(__buf=0x00007f45a99f22d8)::$_0>(DB::KeeperDispatcher::initialize(Poco::Util::AbstractConfiguration const&, bool, bool, std::__1::shared_ptr<DB::Macros const> const&)::$_0&&)::'lambda'(), void ()> >(std::__1::__function::__policy_storage const*) at function.h:716:16
    clickhouse`std::__1::__function::__policy_func<void ()>::operator(this=0x00007f45a99f22d8)[abi:v15000]() const at function.h:848:16
    clickhouse`std::__1::function<void ()>::operator(this=0x00007f45a99f22d8)() const at function.h:1187:12
    clickhouse`ThreadPoolImpl<std::__1::thread>::worker(this=0x00007f475c8229c0, thread_it=std::__1::list<std::__1::thread, std::__1::allocator<std::__1::thread> >::iterator @ 0x00007f45a99f23a8) at ThreadPool.cpp:427:13
    clickhouse`void ThreadPoolImpl<std::__1::thread>::scheduleImpl<void>(this=0x00007f45cea0c8a8)>, Priority, std::__1::optional<unsigned long>, bool)::'lambda0'()::operator()() const at ThreadPool.cpp:180:73
    clickhouse`decltype(__f=0x00007f45cea0c8a8)()) std::__1::__invoke[abi:v15000]<void ThreadPoolImpl<std::__1::thread>::scheduleImpl<void>(std::__1::function<void ()>, Priority, std::__1::optional<unsigned long>, bool)::'lambda0'()>(void&&) at invoke.h:394:23
    clickhouse`void std::__1::__thread_execute[abi:v15000]<std::__1::unique_ptr<std::__1::__thread_struct, std::__1::default_delete<std::__1::__thread_struct> >, void ThreadPoolImpl<std::__1::thread>::scheduleImpl<void>(std::__1::function<void ()>, Priority, std::__1::optional<unsigned long>, bool)::'lambda0'()>(__t=size=2, (null)=__tuple_indices<> @ 0x00007f45a99f2408)>, Priority, std::__1::optional<unsigned long>, bool)::'lambda0'()>&, std::__1::__tuple_indices<>) at thread:284:5
    clickhouse`void* std::__1::__thread_proxy[abi:v15000]<std::__1::tuple<std::__1::unique_ptr<std::__1::__thread_struct, std::__1::default_delete<std::__1::__thread_struct> >, void ThreadPoolImpl<std::__1::thread>::scheduleImpl<void>(std::__1::function<void ()>, Priority, std::__1::optional<unsigned long>, bool)::'lambda0'()> >(__vp=0x00007f45cea0c8a0) at thread:295:5
    libc.so.6`___lldb_unnamed_symbol3481 + 755
    libc.so.6`___lldb_unnamed_symbol3865 + 11

This should fix the problem with
test_keeper_incorrect_config/test.py::test_invalid_configs test.

CI: https://s3.amazonaws.com/clickhouse-test-reports/52717/72b1052f7c2d453308262924e767ab8dc2206933/integration_tests__asan__[1_6].html
Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2023-12-09 17:54:16 +01:00
Azat Khuzhin
55d5a3affa Correctly wait background threads
There are some places that make it possible to trigger use-after-free
from threads because some global variable had been destroyed before, for
example some logger.

Signed-off-by: Azat Khuzhin <a.khuzhin@semrush.com>
2023-12-09 17:54:16 +01:00
Han Fei
881e58dfcb support soft limit ratio 2023-12-08 21:48:27 +01:00
Alexey Milovidov
0b597eadf8 Clarify the usage of prefer_not_to_merge option 2023-12-06 12:11:03 +01:00
jsc0218
cdd5280272
Merge pull request #57387 from evillique/better-disks
Initialize only required disks in clickhouse-disks
2023-12-05 13:59:04 -05:00
Alexey Milovidov
cbea3545aa
Revert "add function getClientHTTPHeader" 2023-12-05 13:34:34 +03:00
Sergei Trifonov
c073d03a20
Merge pull request #57236 from ClickHouse/multiline-charts
Support multiple lines on dashboard charts
2023-12-04 17:35:31 +01:00
serxa
1c650f380a remove debug print 2023-12-04 16:26:39 +00:00
Alexey Milovidov
02439eee58
Merge pull request #49351 from ClickHouse/async-loader-integration
Asynchronous loading of tables
2023-12-04 17:16:12 +01:00
Nikolay Degterinsky
93995f6c06 Fix test 2023-12-02 23:59:46 +00:00
Nikita Taranov
a32c5e976a
Don't print server revision in client on connect (#57435) 2023-12-02 16:32:34 +01:00
Nikolay Degterinsky
e8203f8a76 Initialize only required disks 2023-11-30 03:09:55 +00:00
Antonio Andelic
93360e4e2c
Merge pull request #57299 from ClickHouse/keeper-changelog-snapshot-fix
Keeper fix for changelog and snapshots
2023-11-29 08:31:48 +01:00
János Benjamin Antal
4c2451f837
Merge pull request #56741 from ClickHouse/well-known-protos 2023-11-28 15:21:15 +01:00
Antonio Andelic
a61f32867e
Merge pull request #56488 from lingtaolf/feature/getHTTPHeader
add function getClientHTTPHeader
2023-11-28 11:41:56 +01:00
Antonio Andelic
889c53eca0 Correctly handle unknown changelog versions and snapshot deleting 2023-11-28 09:47:49 +00:00
serxa
0d39de9ee9 indicate that legend is not full by "..." 2023-11-28 01:18:41 +00:00
serxa
4979925095 fix legend + add string-labeled format 2023-11-28 00:36:38 +00:00
János Benjamin Antal
adf21e37c4 Make tests use the installed protobuf files 2023-11-27 13:39:28 +00:00
János Benjamin Antal
4bfbbfbd75 Fix proto file installation 2023-11-27 13:39:01 +00:00
serxa
62831d235f support multiple lines on dashboard charts 2023-11-26 23:28:11 +00:00
Vitaly Baranov
5769a88b92
Merge pull request #57133 from vitlibar/change-default-for-wait_dictionaries_load_at_startup
Change the default for wait_dictionaries_load_at_startup to true
2023-11-24 17:09:05 +01:00
alesapin
a3c2896e9d
Merge pull request #56846 from aalexfvk/add_cancellation_point_for_moving_background_operation
Add cancellation hook for moving background operation
2023-11-24 12:12:47 +01:00
Kseniia Sumarokova
650c1e959d
Merge pull request #57141 from ClickHouse/remove-test-case
Follow up to #56541
2023-11-24 11:45:29 +01:00
Alexey Milovidov
98cddf5312
Merge pull request #57153 from ClickHouse/better-server-exit-code
Better server exit code
2023-11-23 23:55:28 +01:00
Igor Nikonov
0434044a7e
Merge pull request #57054 from ClickHouse/pr-cleanup-narrow-dependency
Parallel replicas: cleanup, narrow dependency
2023-11-23 20:12:37 +01:00
Sergei Trifonov
9436ae6d7d
Merge pull request #56771 from ClickHouse/dashboards-table
Load queries for dashboard page from new `system.dashboards` table
2023-11-23 19:07:52 +01:00
Alexey Milovidov
b62e9d3a1a Address review comments 2023-11-23 18:13:12 +01:00
Alexey Milovidov
137bf2cf9c Whitespaces 2023-11-23 17:51:55 +01:00
Alexey Milovidov
bd4ffec361 Better exit codes from the server 2023-11-23 17:00:57 +01:00
Alexey Milovidov
5abed8f8a8 Better exit codes from the server 2023-11-23 17:00:02 +01:00
Vitaly Baranov
5f461ff780 Change the default for "wait_dictionaries_load_at_startup" to true,
and use this setting only if "dictionaries_lazy_load" is false.
2023-11-23 14:45:42 +01:00
Aleksei Filatov
1a03f5f7f4 Merge remote-tracking branch 'upstream/master' into add_cancellation_point_for_moving_background_operation 2023-11-23 16:43:33 +03:00
kssenii
ff3224e7f7 Load named collections in advance 2023-11-23 13:22:06 +01:00
serxa
663c8cd944 Merge branch 'master' into dashboards-table 2023-11-23 10:29:48 +00:00
serxa
0fc25fbda8 Merge branch 'master' into async-loader-integration 2023-11-23 09:32:43 +00:00
János Benjamin Antal
ab935e3dd7 Use the google proto files when importing protobuf schemas 2023-11-22 12:39:41 +00:00
János Benjamin Antal
6f652133a8 Install well-known protobuf types 2023-11-22 12:39:29 +00:00
Kseniia Sumarokova
1dfafcfe5a
Merge branch 'master' into add-system-s3-queue-log-to-config 2023-11-21 21:31:35 +01:00
vdimir
a139ae97eb
Merge pull request #52918 from ClickHouse/vdimir/s3_blob_log
Add system table with blob storage operations log
2023-11-21 17:40:42 +01:00
Igor Nikonov
63a1625b77 Merge remote-tracking branch 'origin/master' into pr-cleanup-narrow-dependency 2023-11-21 16:05:48 +00:00
Igor Nikonov
ce98dfb251 Settings pointer to reference 2023-11-21 16:04:54 +00:00
凌涛
e86aa433c5
Merge branch 'master' into feature/getHTTPHeader 2023-11-21 19:48:47 +08:00
kssenii
ad9d4a9bb8 Add system.s3queue_log to default config 2023-11-21 11:53:43 +01:00