[
https://issues.apache.org/jira/browse/SPARK-6952?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14499743#comment-14499743
]
Sean Owen commented on SPARK-6952:
----------------------------------
It's too late for 1.3.1, but not too late to put it in branch 1.3 for 1.3.2. I
think that's OK. I should also then back-port SPARK-5825, which is the same
change for another instance in the file.
> spark-daemon.sh PID reuse check fails on long classpath
> -------------------------------------------------------
>
> Key: SPARK-6952
> URL: https://issues.apache.org/jira/browse/SPARK-6952
> Project: Spark
> Issue Type: Bug
> Components: Deploy
> Affects Versions: 1.3.0
> Reporter: Punya Biswal
> Assignee: Punya Biswal
> Priority: Minor
> Fix For: 1.4.0
>
>
> {{sbin/spark-daemon.sh}} uses {{ps -p "$TARGET_PID" -o args=}} to figure out
> whether the process running with the expected PID is actually a Spark daemon.
> When running with a large classpath, the output of {{ps}} gets truncated and
> the check fails spuriously.
> I think we should weaken the check to see if it's a java command (which is
> something we do in other parts of the script) rather than looking for the
> specific main class name. This means that SPARK-4832 might happen under a
> slightly broader range of circumstances (a *java* program happened to reuse
> the same PID), but it seems worthwhile compared to failing consistently with
> a large classpath.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
