[
https://issues.apache.org/jira/browse/FLINK-16215?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17045565#comment-17045565
]
YufeiLiu edited comment on FLINK-16215 at 2/26/20 2:08 PM:
-----------------------------------------------------------
[~xintongsong] I understand your concern, so we can't know how many slots will
be recovered when JM failover. I think we can keep current behavior until
there is timeout check with worker startup.
I came up with this because the issue we discuss before
[FLINK-15959|https://issues.apache.org/jira/browse/FLINK-15959], it's hardly to
know exactly how many slots are missing at startup.
was (Author: liuyufei):
[~xintongsong] I understand your concern, so we can't know how many slots will
be recovered when JM failover. I came up with this because the issue we discuss
before [FLINK-15959|https://issues.apache.org/jira/browse/FLINK-15959], it's
hardly to know exactly how many slots are missing at startup.
> Start redundant TaskExecutor when JM failed
> -------------------------------------------
>
> Key: FLINK-16215
> URL: https://issues.apache.org/jira/browse/FLINK-16215
> Project: Flink
> Issue Type: Bug
> Components: Runtime / Coordination
> Affects Versions: 1.10.0
> Reporter: YufeiLiu
> Priority: Major
>
> TaskExecutor will reconnect to the new ResourceManager leader when JM failed,
> and JobMaster will restart and reschedule job. If job slot request arrive
> earlier than TM registration, RM will start new workers rather than reuse the
> existing TMs.
> It‘s hard to reproduce becasue TM registration usually come first, and
> timeout check will stop redundant TMs.
> But I think it would be better if we make the {{recoverWokerNode}} to
> interface, and put recovered slots in {{pendingSlots}} wait for TM
> reconnection.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
