[
https://issues.apache.org/jira/browse/FLINK-13702?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16962953#comment-16962953
]
Jark Wu commented on FLINK-13702:
---------------------------------
Hi [~dwysakowicz], thanks for looking into this. However, I think they will be
called by different threads. I think the root cause is the same with
FLINK-13740.
The reason is that the async snapshot thread (e.g. snapshot the heap
CopyOnWriteStateMap to disk) and the process element thread are concurrent. The
former one will call {{TypeSerializer#serialize()}}, and the latter one will
call {{TypeSerializer#duplicate()}} (when the state entry is in snapshoting)
on the same object. I think that's why {{SerializerTestBase#testDuplicate()}}
is added, and that's why {{LazyBinaryFormat#ensureMaterialized}} will be called
concurrently.
So a simple way to fix this is adding synchronize to {{ensureMaterialized}},
but I'm not sure how much the performance impact is.
> BaseMapSerializerTest.testDuplicate fails on Travis
> ---------------------------------------------------
>
> Key: FLINK-13702
> URL: https://issues.apache.org/jira/browse/FLINK-13702
> Project: Flink
> Issue Type: Bug
> Components: Table SQL / Planner
> Affects Versions: 1.10.0
> Reporter: Till Rohrmann
> Assignee: Dawid Wysakowicz
> Priority: Critical
> Labels: test-stability
>
> The {{BaseMapSerializerTest.testDuplicate}} fails on Travis with an
> {{java.lang.IndexOutOfBoundsException}}.
> https://api.travis-ci.org/v3/job/570973199/log.txt
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
