Add logs_to_keep to replicated database settings

[Khatskevich]

Changelog category (leave one):

• Improvement

Changelog entry (a user-readable short description of the changes that goes into CHANGELOG.md):

Added a server setting, logs_to_keep to database replicated settings, that allows changing the default logs_to_keep parameter for replicated databases. Lower values reduce the number of ZNodes (especially if there are many databases), while higher values allow a missing replica to catch up after a longer period of time.

Documentation entry for user-facing changes

• Documentation is written (mandatory for new features)
  (short docs are written directly into src/Core/ServerSettings.cpp

[clickhouse-gh]

Workflow [PR], commit [8786630]

[tuanpach]

We must declare the setting before using it:

ClickHouse/src/Databases/DatabaseReplicated.cpp

Lines 75 to 79 in cbab61c

	namespace ServerSetting
	{
	extern const ServerSettingsBool database_replicated_allow_detach_permanently;
	extern const ServerSettingsUInt32 max_database_replicated_create_table_thread_pool_size;
	}

ClickHouse/src/Databases/DatabaseReplicated.cpp

Line 1556 in cbab61c

auto allow_concurrent_table_creation = getContext()->getServerSettings()[ServerSetting::max_database_replicated_create_table_thread_pool_size] != 1;

[Khatskevich]

Thanks. Declared.

[tuanpach]

We should move this setting to DatabaseReplicatedSettings

https://github.com/ClickHouse/ClickHouse/blob/master/src/Databases/DatabaseReplicatedSettings.cpp#L10

[tuanpach]

• 01825_new_type_json_ghdata_insert_select
  • 01825_new_type_json_ghdata_insert_select is flaky #84577
• 03273_ghdata_json_to_json_alter
  • Test 03273_ghdata_json_to_json_alter is flaky #85148
• Server died:
  • Logical error: Expected JOIN table expression to be table, table function, query or union node #81144
• Logical error thrown
  • DatabaseCatalog: Cannot drop table test_11.data_rmt (c58a0c3a-f677-4b63-836d-17f3aa145faa). Will retry later.: Code: 49. DB::Exception: There is a race condition between creation and removal of replicated table. It's a bug #85771

[tuanpach]

@Khatskevich, I made some changes. Please take a look. Thanks

[Khatskevich]

@tuanpach looks good, thank you!
The only possibility to improve it would be to make the logs_to_keep variable dynamic and adjustable after a database creation, but to me it looks like overkill. It makes sense to leave it as is.

[tuanpach]

• 01825_new_type_json_ghdata_insert_select

  • 01825_new_type_json_ghdata_insert_select is flaky #84577

• Stateless tests (amd_tsan, s3 storage, parallel): 03273_ghdata_json_to_json_alter

  • Test 03273_ghdata_json_to_json_alter is flaky #85148

• Stress test (amd_msan): Server died

  • Assertion when joining non-mertgree with parallel replicas. Logical error: 'Expected JOIN table expression to be table, table function, query or union node. Actual ANY LEFT JOIN ... ON k1 = mem.k' #63984

• Upgrade check (amd_debug): Logical error thrown (see clickhouse-server.log or logical_errors.txt)

  • DatabaseCatalog: Cannot drop table test_11.data_rmt (c58a0c3a-f677-4b63-836d-17f3aa145faa). Will retry later.: Code: 49. DB::Exception: There is a race condition between creation and removal of replicated table. It's a bug #85771

[tuanpach]

• test_memory_limit/test.py::test_multiple_queries

  • test_memory_limit/test.py::test_multiple_queries flakes #72458

• 03595_alter_drop_column_comment_if_exists

  • Fix broken test 03595_alter_drop_column_comment_if_exists #86001

• Logical error: 'Arguments of 'plus' have incorrect data types: '__table1.x' of type 'UInt32', '__table1.d' of type 'Date32''.

  • Logical Error: Arguments of 'minus' have incorrect data types #71415