[
https://issues.apache.org/jira/browse/FLINK-25055?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17449622#comment-17449622
]
Till Rohrmann commented on FLINK-25055:
---------------------------------------
Hi [~zjureel], are you referring to the case that a pipelined region is
deployed, let's say consisting of A -> B (with A and B being operators), and
that B gets deployed before A has registered its partition on the {{TM}}
because there was some race condition?
If we know that a partition will be registered on a given {{TM}}, then we could
indeed register a callback that is being called once this happens. This could
avoid having to retry with an exponential retry mechanism. On both sides, we
should probably register timeouts to remove the callback and fail the
initialization of the {{Task}}, respectively. So I believe that your idea could
improve Flink behaviour.
> Support listen and notify mechanism for PartitionRequest
> --------------------------------------------------------
>
> Key: FLINK-25055
> URL: https://issues.apache.org/jira/browse/FLINK-25055
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Network
> Affects Versions: 1.14.0, 1.12.5, 1.13.3
> Reporter: Shammon
> Priority: Major
>
> We submit batch jobs to flink session cluster with eager scheduler for olap.
> JM deploys subtasks to TaskManager independently, and the downstream subtasks
> may start before the upstream ones are ready. The downstream subtask sends
> PartitionRequest to upstream ones, and may receive PartitionNotFoundException
> from them. Then it will retry to send PartitionRequest after a few ms until
> timeout.
> The current approach raises two problems. First, there will be too many retry
> PartitionRequest messages. Each downstream subtask will send PartitionRequest
> to all its upstream subtasks and the total number of messages will be O(N*N),
> where N is the parallelism of subtasks. Secondly, the interval between
> polling retries will increase the delay for upstream and downstream tasks to
> confirm PartitionRequest.
> We want to support listen and notify mechanism for PartitionRequest when the
> job needs no failover. Upstream TaskManager will add the PartitionRequest to
> a listen list with a timeout checker, and notify the request when the task
> register its partition in the TaskManager.
> [~nkubicek] I noticed that your scenario of using flink is similar to ours.
> What do you think? And hope to hear from you [~trohrmann] THX
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
