[
https://issues.apache.org/jira/browse/FLINK-8922?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16400580#comment-16400580
]
Stephan Ewen commented on FLINK-8922:
-------------------------------------
If I see this correctly, we still assume it is related to object release,
because it occurs only on Travis (rarely on developer laptops) related to GCs
happening.
Another think we can try is to have a {{private static final WriteOptions}}
object in the {{RocksDBKeyedStateBackend}} that simply never gets released. We
need to have that object's initialization after the {{System.loadLibrary()}}
call, though. That could help.
If that does not help, it might not be related to GC / object release after
all, and really be a WAL bug in RocksDB.
> Revert FLINK-8859 because it causes segfaults in testing
> --------------------------------------------------------
>
> Key: FLINK-8922
> URL: https://issues.apache.org/jira/browse/FLINK-8922
> Project: Flink
> Issue Type: Bug
> Components: State Backends, Checkpointing
> Affects Versions: 1.5.0
> Reporter: Stefan Richter
> Assignee: Stefan Richter
> Priority: Major
> Fix For: 1.5.0
>
>
> We need to revert FLINK-8859 because it causes problems with RocksDB that
> make our automated tests fail on Travis. The change looks actually good and
> it is currently unclear why this can introduce such a problem. This might
> also be a Rocks in RocksDB. Nevertheless, for the sake of a proper release
> testing, we should revert the change for now.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
