[
https://issues.apache.org/jira/browse/FLINK-32746?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
xiangyu feng updated FLINK-32746:
---------------------------------
Description:
In a OLAP session cluster, a TM need to frequently create new classloaders and
generate new classes. These classes will be accumulated in metaspace. When
metaspace data usage reaches a threshold, a FullGC with a long time
Stop-the-World will be triggered. Currently, both SerialGC, ParallelGC and G1GC
are doing Stop-the-World class unloading. Only ZGC supports concurrent class
unload, see more in
[https://bugs.openjdk.org/browse/JDK-8218905|https://bugs.openjdk.org/browse/JDK-8218905).].
In our scenario, a class unloading for a 2GB metaspace with 5million classes
will stop the application more than 40 seconds. After switch to ZGC, the
maximum STW of the application has been reduced to less than 10ms.
was:
In a OLAP session cluster, a TM need to frequently create new classloaders and
generate new classes. These classes will be accumulated in metaspace. When
metaspace data usage reaches a threshold, a FullGC with a long time
Stop-the-World will be triggered. Currently, both SerialGC, ParallelGC and G1GC
are doing Stop-the-World class unloading. Only ZGC support concurrent class
unload
> Enable ZGC in JDK17 to solve long time class unloading STW during fullgc
> ------------------------------------------------------------------------
>
> Key: FLINK-32746
> URL: https://issues.apache.org/jira/browse/FLINK-32746
> Project: Flink
> Issue Type: Sub-task
> Components: Table SQL / Runtime
> Reporter: xiangyu feng
> Priority: Major
>
> In a OLAP session cluster, a TM need to frequently create new classloaders
> and generate new classes. These classes will be accumulated in metaspace.
> When metaspace data usage reaches a threshold, a FullGC with a long time
> Stop-the-World will be triggered. Currently, both SerialGC, ParallelGC and
> G1GC are doing Stop-the-World class unloading. Only ZGC supports concurrent
> class unload, see more in
> [https://bugs.openjdk.org/browse/JDK-8218905|https://bugs.openjdk.org/browse/JDK-8218905).].
>
> In our scenario, a class unloading for a 2GB metaspace with 5million classes
> will stop the application more than 40 seconds. After switch to ZGC, the
> maximum STW of the application has been reduced to less than 10ms.
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
