SEPURI-SAI-KRISHNA commented on PR #12495:
URL: https://github.com/apache/seatunnel/pull/12495#issuecomment-6074467190
Build triage for `4b2cd492e`. Four failures and one timeout, none
attributable to this change.
Run `37759695509`: 78 success, 4 failure, 1 cancelled, 11 skipped. Each
verdict is from that job's own test-report artifact.
| job | failure | status |
| --- | --- | --- |
| `engine-v2-it (11)` |
`SplitClusterFaultToleranceIT.testStreamJobCancelResolvesWhenWorkerCrashesBeforeCancelAck`,
11 tests 1 error | #12353, with #12311 as the fix |
| `all-connectors-it-7 (11)` |
`PostgresCDCIT.testPostgresCdcSnapshotOnlyAndCommittedOffsetStartupModes`,
expected `1` but was `0`, 37 tests 1 error | #11847 |
| `rocketmq-connector-it (11)` | `RocketMqIT.testTextFormatSinkRocketMq` on
two invocations, `<10>` vs `<6>` and `<10>` vs `<14>`, 87 tests 2 failures |
**now #12700**, fix in #12701 |
| `mysql-cdc-connector-it (8)` |
`MysqlCDCIT.testMultiTableWithRestore{TestContainer}[1]`, expected `<false>`
but was `<true>`, 48 tests 1 failure | untracked |
| `Build website` | 90.4 min against `timeout-minutes: 90`, recorded as
cancelled | #12614, already in `dev` |
None of the four is a timeout: they ran 91.5, 81.9, 65.4 and 97.3 minutes
against limits of 150, 150, 120 and 120.
**On the RocketMQ one.** `6 + 14 = 20` is exactly two invocations' worth of
rows, so nothing was lost: the read does not wait for the rows it just wrote to
become visible, and because the topic and consumer group are shared across
template invocations the shortfall is inherited by the next one. Filed as
#12700 with the fix in #12701. To be clear, that accounts for **one** of the
four failures above, not the rest.
The `MysqlCDCIT` case is a single occurrence with no issue naming it, so I
am noting rather than filing it.
**Not reachable from this PR.** The diff is seven `src/main` files under
`transform/sql/` plus five test classes in the same module. The jobs that
exercise it all passed:
```text
unit-test success on all 4 legs
transform-v2-it-part-1 success on both legs
transform-v2-it-part-2 success on both legs
```
A rebase onto current `dev` would clear the `Build website` entry outright
via #12614 and give a cleaner matrix to judge the rest against. Happy to do
that.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
