[
https://issues.apache.org/jira/browse/FLINK-8846?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sihua Zhou closed FLINK-8846.
-----------------------------
Resolution: Staged
Fix Version/s: (was: 1.6.0)
Not the right time to do this, we should wait until RocksDB support ingesting
the existing SST into rocksdb or RocksDB has improved the {{SstFileWriter}}'s
performance for java api.
> Introduce `parallel recovery` mode for incremental checkpoint
> -------------------------------------------------------------
>
> Key: FLINK-8846
> URL: https://issues.apache.org/jira/browse/FLINK-8846
> Project: Flink
> Issue Type: Improvement
> Components: State Backends, Checkpointing
> Affects Versions: 1.5.0
> Reporter: Sihua Zhou
> Assignee: Sihua Zhou
> Priority: Major
>
> escription
> Base on {{ingestExternalFile()}} and {{SstFileWriter}} provided by RocksDB,
> we can restore from incremental checkpoint in parallel.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
