Ilya Yatsishin
Ilya Yatsishin
ClickHouse Stateless Tests: #38337 Integration tests #38260 ClickHouse Stress Test: https://github.com/ClickHouse/ClickHouse/commit/299f102ccf92e1ea001ad9483171f75f33cb75bc
@tavplubix it will be easier if https://github.com/ClickHouse/ClickHouse/commit/299f102ccf92e1ea001ad9483171f75f33cb75bc was a PR to master. I could merge it just by setting the label. Now it requires manual merge
OK. Haven't seen this issue. It means current release is incompatible.
@tavplubix https://s3.amazonaws.com/clickhouse-test-reports/38146/1fc97f10cbffda0ae8460c575a4637e45e09cc11/stress_test__address__actions_.html It is reproducing https://github.com/ClickHouse/ClickHouse/issues/37664 Proposed fix is already in 22.6.1 https://github.com/ClickHouse/ClickHouse/pull/37839/commits/2358878c381dcef89a288a414679017dc22a2a86
00984_parser_stack_overflow became flaky in June. Need to check if it the real issue https://play.clickhouse.com/play?user=play#c2VsZWN0IGNoZWNrX25hbWUsIHRlc3RfbmFtZSwgY2hlY2tfc3RhcnRfdGltZSwgdGVzdF9zdGF0dXMsIHJlcG9ydF91cmwsIGNvbW1pdF9zaGEKZnJvbSBjaGVja3MgCndoZXJlIAp0ZXN0X3N0YXR1cyBub3QgaW4gWydPSycsICdTVUNDRVNTJywgJ1NLSVBQRUQnLCAnWEZBSUwnLCAnWEVSUk9SJ10gYW5kIAp0ZXN0X25hbWUgbGlrZSAnMDA5ODRfcGFyc2VyX3N0YWNrX292ZXJmbG93JScgCmFuZCBwdWxsX3JlcXVlc3RfbnVtYmVyID0gMAphbmQgY2hlY2tfc3RhcnRfdGltZSA+PSBub3coKSAtIGludGVydmFsIDMwIGRheQpvcmRlciBieSBjaGVja19zdGFydF90aW1lIGRlc2M=
With tsan 00984_parser_stack_overflow really takes more than 1 minute ``` 2022.06.17 16:58:15.473366 [ 22916 ] {e0efd9c8-eef0-49d4-b283-469b5df2f3a4} executeQuery: (from [::1]:50398) (comment: 00984_parser_stack_overflow.sh) SELECT ([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([([> 2022.06.17 16:59:42.964748 [ 22916 ] {e0efd9c8-eef0-49d4-b283-469b5df2f3a4} executeQuery:...
>There might be conflicts once https://github.com/ClickHouse/ClickHouse/pull/38965 is merged No worries. Your PR if merged will make that one irrelevant.
This PR fixes #39122?
@FrankChen021 will you look at failed tests? Failure in `01526_initial_query_id` looks related and other tests probably failed due to changes in context propagation
I asked other developers to look at it. I just don't know this part.