SEZ9 commented on PR #12595:
URL: https://github.com/apache/seatunnel/pull/12595#issuecomment-6159669970
@SEPURI-SAI-KRISHNA thanks, and you are right on the correction: I
misattributed the JDK 8 check to you. Your disassembly covered 11.0.32, 17.0.20
and 21.0.12 (length folded into the accumulator); the 8u71/8u171 finding
(`if_icmpeq` on the lengths, then `return false`) came from the other review,
not from you. The javadoc wording you describe is the accurate one:
leading-byte property holds on every shape, only the length stays inferable on
8u. One small ask so I can close this out: please confirm that wording is what
is in the pushed head `364514f12c`, or push it if it is still local.
On the invocation mapping, that fully answers my question. `[2]` = Flink
1.20.1, Flink 1.18.0 passing in the same job, and the `Direct buffer memory` ->
`ConnectionInitException` -> `AllNodesFailedException` -> exit 1 chain is a
much better artifact than the `{TestContainer}[2]` signature alone. Thanks for
filing #12757 with both reproductions including `dev` head `b0f6feddf` on run
38340962420, and for flagging the #12732 overlap for reconciliation rather than
asserting it; that is the right way to leave it.
Since #12756 (`07d6f2235a`) landed on `dev` and only touches
`Flink20Container.java` in `seatunnel-e2e-common`, the remaining steps are:
1. Merge latest `dev` into this branch so the Flink 1.20 direct-memory fix
is picked up.
2. Let `Build` run again; the `CassandraIT` failures on the Flink 1.20.1
container should disappear.
3. If `all-connectors-it-6` (`DatabendCDCSinkIT`) is still red afterwards,
that one is independent and a single failed-jobs rerun is enough.
Nothing else outstanding from me on the code; once the javadoc confirmation
and a green run after the `dev` sync are in, I am ready to approve.
<!-- streview-comment:1736 -->
--
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]
