[
https://issues.apache.org/jira/browse/AMBARI-17539?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andrew Onischuk updated AMBARI-17539:
-------------------------------------
Description:
*Reason of memory leak:*
Race condition in subprocess python module.
Due to this race condition at some unlucky cases python garbage collection was
disabled.
This usually happened when running alerts, as a bunch of our alerts run shell
commands and they do it in different threads.
*Fix for the issue:*
Synchronizing subprocess is not the best option. Since some people can still
use it without synchronization not knowing about the issue.
Also synchronizing will provide some unnecessary slowdown. So for this issue
the proposed fix is to monkey patch subprocess.gc.isenabled.
was:
*Reason of memory leak:*
Race condition in subprocess python module.
Due to this race condition at some unlucky cases python garbage collection was
disabled.
This usually happened when running alerts, as a bunch of our alerts run shell
commands and they do it in a different threads.
*Fix for the issue:*
Synchronizing subprocess is not the best option. Since some people can still
use it without synchronization not knowing about the issue.
Also synchronizing will provide some unnecessary slowdown. So for this issue
the proposed fix is to monkey patch subprocess.gc.isenabled.
> Ambari Agent memory Leak fix.
> -----------------------------
>
> Key: AMBARI-17539
> URL: https://issues.apache.org/jira/browse/AMBARI-17539
> Project: Ambari
> Issue Type: Bug
> Reporter: Andrew Onischuk
> Assignee: Andrew Onischuk
> Fix For: 2.4.0
>
> Attachments: AMBARI-17539.patch, AMBARI-17539.patch
>
>
> *Reason of memory leak:*
> Race condition in subprocess python module.
> Due to this race condition at some unlucky cases python garbage collection
> was disabled.
> This usually happened when running alerts, as a bunch of our alerts run shell
> commands and they do it in different threads.
> *Fix for the issue:*
> Synchronizing subprocess is not the best option. Since some people can still
> use it without synchronization not knowing about the issue.
> Also synchronizing will provide some unnecessary slowdown. So for this issue
> the proposed fix is to monkey patch subprocess.gc.isenabled.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
