[
https://issues.apache.org/jira/browse/BEAM-5397?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17238650#comment-17238650
]
Huang Xingbo commented on BEAM-5397:
------------------------------------
We encountered the same problem in pyflink
https://issues.apache.org/jira/browse/FLINK-20284. After investigation, I found
that the reason for the problem is that Grpc releases the related resources of
Netty Server by schedule tasks started by 1s.
https://github.com/grpc/grpc-java/blob/master/core/src/main/java/io/grpc/internal/SharedResourceHolder.java#L150.
After 1s, when the schedule Task starts to reclaim resources, the related
classes have been unloaded by the UserClassLoader, so ClassNotFoundError is
thrown.
> Flink portable runner GRPC cleanup failure after user class loader was removed
> ------------------------------------------------------------------------------
>
> Key: BEAM-5397
> URL: https://issues.apache.org/jira/browse/BEAM-5397
> Project: Beam
> Issue Type: Improvement
> Components: runner-flink
> Affects Versions: 2.8.0
> Reporter: Thomas Weise
> Priority: P3
> Labels: portability
>
> Looks like another attempt to perform cleanup after close.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
