[GitHub] [flink] pgaref commented on a diff in pull request #22506: [FLINK-31890][runtime] Introduce JobMaster per-task failure enrichment/labeling

[via GitHub]

pgaref commented on code in PR #22506:
URL: https://github.com/apache/flink/pull/22506#discussion_r1185674060


##########
flink-runtime/src/main/java/org/apache/flink/runtime/executiongraph/Execution.java:
##########
@@ -778,7 +778,7 @@ private static PartitionInfo createFinishedPartitionInfo(
      */
     @Override
     public void fail(Throwable t) {
-        processFail(t, true);
+        processFail(t, true, Collections.emptyMap());

Review Comment:
   Again this is covered with the next PR: 
https://github.com/apache/flink/pull/22511/files#diff-f9e390431932be9a4505e581632815b81c1f0520e65cec7614ea4b90c8a52afcR781-R782
   
   But there are still corner cases within the ExecutionGraph that are not 
easily covered, e.g., 
[deploy](https://github.com/pgaref/flink/blob/41728f1bae85ce19f843fc18d0d80ddfe75af6b9/flink-runtime/src/main/java/org/apache/flink/runtime/executiongraph/Execution.java#L608),
 
[triggerTaskFailover](https://github.com/apache/flink/pull/22511/files#diff-c7f8fc6bbf59b935414529c76720dcafce6105e16783b918af6b06a10e247ab2R103-R104),
 
[UpdatePartitionInfo](https://github.com/apache/flink/pull/22511/files#diff-f9e390431932be9a4505e581632815b81c1f0520e65cec7614ea4b90c8a52afcR1378)
 
   
   The intuition here was that all these failures are internal and could maybe 
skip labeling -- Davids suggestion following all the `fromSchedulerNg  == 
false` paths also seems to cover this --  but happy to discuss other thoughts?



-- 
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: issues-unsubscr...@flink.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org