[
https://issues.apache.org/jira/browse/FLINK-6504?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16007598#comment-16007598
]
ASF GitHub Bot commented on FLINK-6504:
---------------------------------------
Github user shixiaogang commented on a diff in the pull request:
https://github.com/apache/flink/pull/3859#discussion_r116148624
--- Diff:
flink-contrib/flink-statebackend-rocksdb/src/main/java/org/apache/flink/contrib/streaming/state/RocksDBKeyedStateBackend.java
---
@@ -911,9 +915,11 @@ void releaseResources(boolean canceled) {
if (canceled) {
List<StateObject> statesToDiscard = new
ArrayList<>();
- statesToDiscard.add(metaStateHandle);
- statesToDiscard.addAll(miscFiles.values());
- statesToDiscard.addAll(newSstFiles.values());
+ synchronized (this) {
--- End diff --
Yes, i agree. The key point here is to make sure the stopping of the
materialization thread. Synchronization does little help here. So i prefer to
remove synchronization here, what do you think?
> Lack of synchronization on materializedSstFiles in RocksDBKEyedStateBackend
> ---------------------------------------------------------------------------
>
> Key: FLINK-6504
> URL: https://issues.apache.org/jira/browse/FLINK-6504
> Project: Flink
> Issue Type: Sub-task
> Components: State Backends, Checkpointing
> Affects Versions: 1.3.0
> Reporter: Stefan Richter
> Assignee: Xiaogang Shi
> Priority: Blocker
> Fix For: 1.3.0
>
>
> Concurrent checkpoints could access `materializedSstFiles` in the
> `RocksDBStateBackend` concurrently. This should be avoided.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
