[
https://issues.apache.org/jira/browse/FLINK-9416?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16489111#comment-16489111
]
ASF GitHub Bot commented on FLINK-9416:
---------------------------------------
Github user tillrohrmann commented on a diff in the pull request:
https://github.com/apache/flink/pull/6069#discussion_r190609755
--- Diff:
flink-clients/src/main/java/org/apache/flink/client/program/rest/RestClusterClient.java
---
@@ -314,7 +319,8 @@ public JobSubmissionResult submitJob(JobGraph jobGraph,
ClassLoader classLoader)
log.info("Requesting blob server port.");
CompletableFuture<BlobServerPortResponseBody> portFuture =
sendRequest(
- BlobServerPortHeaders.getInstance());
+ BlobServerPortHeaders.getInstance(),
+ EmptyMessageParameters.getInstance());
--- End diff --
Good point. Will change it.
> Make job submission retriable operation in case of a ongoing leader election
> ----------------------------------------------------------------------------
>
> Key: FLINK-9416
> URL: https://issues.apache.org/jira/browse/FLINK-9416
> Project: Flink
> Issue Type: Bug
> Components: Client
> Affects Versions: 1.5.0, 1.6.0
> Reporter: Till Rohrmann
> Assignee: Till Rohrmann
> Priority: Critical
> Fix For: 1.6.0, 1.5.1
>
>
> When starting a session cluster, it can happen that the job submission fails
> if the REST server endpoint has already gained leadership but if the
> leadership election for the {{Dispatcher}} is still ongoing. In such a case,
> we receive a error response saying that the leader election is still ongoing
> and fail the job submission. I think it would be nicer to also make the
> submission step a retriable operation in order to avoid this race condition.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
