[ https://issues.apache.org/jira/browse/FLINK-8699?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16417522#comment-16417522 ]
ASF GitHub Bot commented on FLINK-8699: --------------------------------------- Github user tillrohrmann commented on the issue: https://github.com/apache/flink/pull/5705 Hi @sihuazhou, Stefan is currently sick but he should come back in a bit. Since he already reviewed parts of the PR, I would suggest to wait for his return. If he should not be back by next week, I'll take over. > Fix concurrency problem in rocksdb full checkpoint > -------------------------------------------------- > > Key: FLINK-8699 > URL: https://issues.apache.org/jira/browse/FLINK-8699 > Project: Flink > Issue Type: Bug > Components: State Backends, Checkpointing > Affects Versions: 1.5.0 > Reporter: Sihua Zhou > Assignee: Sihua Zhou > Priority: Blocker > Fix For: 1.5.0 > > > In full checkpoint, `kvStateInformation` is not a copied object and it can be > changed when writeKVStateMetaData() is invoking ... This can lead to > problematic, which is serious. > {code} > private void writeKVStateMetaData() throws IOException { > // ... > for (Map.Entry<String, Tuple2<ColumnFamilyHandle, > RegisteredKeyedBackendStateMetaInfo<?, ?>>> column : > stateBackend.kvStateInformation.entrySet()) { > } > //... > } > {code} -- This message was sent by Atlassian JIRA (v7.6.3#76005)