[
https://issues.apache.org/jira/browse/FLINK-8845?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16502809#comment-16502809
]
Sihua Zhou commented on FLINK-8845:
-----------------------------------
Hi [~noliran] Yes, this is on purpose. In fact, we
tried([FLINK-8859|https://issues.apache.org/jira/browse/FLINK-8859]) to disable
WAL in RocksDBKeyedStateBackend when restoring the backend, but it will cause
segfaults on
travis([FLINK-8882|https://issues.apache.org/jira/browse/FLINK-8882]), and the
reason why it caused the segfaults is still not clear, so we
reverted([8922|https://issues.apache.org/jira/browse/FLINK-8922]) it in the
end.
> Use WriteBatch to improve performance for recovery in RocksDB backend
> ---------------------------------------------------------------------
>
> Key: FLINK-8845
> URL: https://issues.apache.org/jira/browse/FLINK-8845
> Project: Flink
> Issue Type: Improvement
> Components: State Backends, Checkpointing
> Affects Versions: 1.5.0
> Reporter: Sihua Zhou
> Assignee: Sihua Zhou
> Priority: Major
> Fix For: 1.5.0
>
>
> Base on {{WriteBatch}} we could get 30% ~ 50% performance lift when loading
> data into RocksDB.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
