[jira] [Commented] (SPARK-1086) Some corner case during HA master switching?

["Apache Spark (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-1086?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14964215#comment-14964215 ] 

Apache Spark commented on SPARK-1086:
-------------------------------------

User 'JihongMA' has created a pull request for this issue:
https://github.com/apache/spark/pull/9172

> Some corner case during HA master switching?
> --------------------------------------------
>
>                 Key: SPARK-1086
>                 URL: https://issues.apache.org/jira/browse/SPARK-1086
>             Project: Spark
>          Issue Type: Bug
>          Components: Deploy
>    Affects Versions: 0.9.0
>            Reporter: Raymond Valencia
>
> In spark standalone HA mode.
> It seems to me that when current master is down, and the new master is not elected yet, at this time if the worker is sending executor update or driver update message. this message will be lost. Though executor status is rebuild after new master is recovered, thus the status might be alright. While the driver might be relaunched even it actually finished successfully, this might be a problem? And then the message send to AppClientListener might be lost and never been regenerated thus lead to some problem?



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org