[jira] [Resolved] (IMPALA-12643) resolve_minidumps.py's minidump_stackwalk call hits OOM on Centos7

["Joe McDonnell (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/IMPALA-12643?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Joe McDonnell resolved IMPALA-12643.
------------------------------------
    Fix Version/s: Impala 4.4.0
       Resolution: Fixed

> resolve_minidumps.py's minidump_stackwalk call hits OOM on Centos7
> ------------------------------------------------------------------
>
>                 Key: IMPALA-12643
>                 URL: https://issues.apache.org/jira/browse/IMPALA-12643
>             Project: IMPALA
>          Issue Type: Bug
>          Components: Infrastructure
>    Affects Versions: Impala 4.4.0
>            Reporter: Joe McDonnell
>            Assignee: Joe McDonnell
>            Priority: Major
>             Fix For: Impala 4.4.0
>
>
> When looking into fixing IMPALA-11909 to enable running resolve_minidumps.py from finalize.sh, testing on Centos 7 showed that the minidump_stackwalk invocation from resolve_minidumps.py went haywire and used all the system's memory.
> resolve_minidumps.py is unique in processing all applicable system libraries. If some system library had corrupt symbols, it could cause minidump_stackwalk to get into an infinite loop.
> We should track down the offending library, but we should also run minidump_stackwalk with a limit on its memory usage to avoid using all the system memory.
>  



--
This message was sent by Atlassian Jira
(v8.20.10#820010)