tillrohrmann commented on a change in pull request #14686:
URL: https://github.com/apache/flink/pull/14686#discussion_r560025485
##########
File path:
flink-runtime/src/main/java/org/apache/flink/runtime/scheduler/DefaultScheduler.java
##########
@@ -270,9 +270,22 @@ private void restartTasksWithDelay(final
FailureHandlingResult failureHandlingRe
delayExecutor.schedule(
() ->
FutureUtils.assertNoException(
- cancelFuture.thenRunAsync(
- restartTasks(executionVertexVersions,
globalRecovery),
- getMainThreadExecutor())),
+ cancelFuture
+ .thenRunAsync(
+ restartTasks(
+
executionVertexVersions, globalRecovery),
+ getMainThreadExecutor())
+ .thenRunAsync(
+ () ->
+ archiveExceptions(
+
failureHandlingResult.getError(),
+
executionVertexVersions.stream()
+ .map(
+
ExecutionVertexVersion
+
::getExecutionVertexId)
+
.collect(
+
Collectors
+
.toList())))),
Review comment:
No, actually I think a better way would be to ask the `Executions` which
are reset whether they have failed or not and if they have to record the
`Execution.failureCause`. For the timestamp you can check whether taking the
state transition timestamp to the `FAILED` state is good enough.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
