[
https://issues.apache.org/jira/browse/FLINK-3396?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15145312#comment-15145312
]
ASF GitHub Bot commented on FLINK-3396:
---------------------------------------
GitHub user uce opened a pull request:
https://github.com/apache/flink/pull/1633
[FLINK-3396] [runtime] Fail job submission after state restore failure
If state restore fails during job graph submission, the submitting client
never gets notified about the submission failure. For detached submissions,
this results in a time out rather than the actual failure cause. The actual
failure can only be seen in the logs.
This PR changes the behaviour to fail the submission.
The failure cause is wrapped in a `SuppressRestartsException` in order to
prevent the restart strategy from kicking in. We call
`ExecutionGraph#fail(Throwable)` in the first place in order to have proper
clean up.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/uce/flink 3396-state_restore
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/flink/pull/1633.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #1633
----
commit 982857bc0f5248a694d908141df0330de2befa3c
Author: Ufuk Celebi <[email protected]>
Date: 2016-02-12T20:59:58Z
[FLINK-3396] [runtime] Fail job submission after state restore failure
Problem: If state restore fails during job graph submission, the submitting
client never gets notified about the submission failure. For detached
submissions, this results in a time out rather than the actual failure
cause.
The actual failure can only be seen in the logs.
Solution: Fail the submission if state restore fails.
commit 9d2f4d88e99c3e50d8c001c8aa2cdb2848c80478
Author: Ufuk Celebi <[email protected]>
Date: 2016-02-12T21:09:29Z
[hotfix] Rename UnrecoverableException to SuppressRestartsException
----
> Job submission Savepoint restore logic flawed
> ---------------------------------------------
>
> Key: FLINK-3396
> URL: https://issues.apache.org/jira/browse/FLINK-3396
> Project: Flink
> Issue Type: Bug
> Reporter: Ufuk Celebi
> Assignee: Ufuk Celebi
> Fix For: 1.0.0
>
>
> When savepoint restoring fails, the thrown Exception fails the execution
> graph, but the client is not informed about the failure.
> The expected behaviour is that the submission should be acked with success or
> failure in any case. With savepoint restore failures, the ack message will be
> skipped.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
