[jira] [Commented] (SPARK-14192) Executor is dead in Driver but alive in AM when driver losts rpc with executor, but executor is alive.

["Shixiong Zhu (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-14192?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15214905#comment-15214905 ] 

Shixiong Zhu commented on SPARK-14192:
--------------------------------------

Could you add the spark version you are using?

> Executor is dead in Driver but alive in AM when driver losts rpc with executor, but executor is alive.
> ------------------------------------------------------------------------------------------------------
>
>                 Key: SPARK-14192
>                 URL: https://issues.apache.org/jira/browse/SPARK-14192
>             Project: Spark
>          Issue Type: Bug
>          Components: Scheduler
>            Reporter: meiyoula
>
> reproduce steps:
> 1. start up spark-sql with: bin/spark-sql --master yarn --num-executors 1
> 2. lost driver and executor connection, but make sure executor is still alive
> 3. execute "show tables", the process will be stuck. Because the driver ask Am to allocate an executor, but AM see the executor is still alive, so don't allocate new one.



--
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