[
https://issues.apache.org/jira/browse/FLINK-10198?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16605516#comment-16605516
]
ASF GitHub Bot commented on FLINK-10198:
----------------------------------------
NicoK commented on issue #6603: [FLINK-10198][state] Set Env object in
DBOptions for RocksDB
URL: https://github.com/apache/flink/pull/6603#issuecomment-419026348
Learning more about this every day, but it looks like the memory options
(writeBufferSize - the mem_table, block cache, etc.) are per column family
anyway and therefore per Flink state and *not* per keyed operator and *not*
globally, even after this PR.
@StefanRRichter to me it looks like this "only" affects the number of
compaction threads (previously `x per keyed operator`, with this PR `x per TM`)
and maybe some configuration properties that are shared with this PR. What else
does this affect?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
> Set Env object in DBOptions for RocksDB
> ---------------------------------------
>
> Key: FLINK-10198
> URL: https://issues.apache.org/jira/browse/FLINK-10198
> Project: Flink
> Issue Type: Improvement
> Affects Versions: 1.7.0
> Reporter: Stefan Richter
> Assignee: Stefan Richter
> Priority: Major
> Labels: pull-request-available
>
> I think we should consider to always set a default environment when we create
> the DBOptions.
> See https://github.com/facebook/rocksdb/wiki/rocksdb-basics:
> *Support for Multiple Embedded Databases in the same process*
> A common use-case for RocksDB is that applications inherently partition their
> data set into logical partitions or shards. This technique benefits
> application load balancing and fast recovery from faults. This means that a
> single server process should be able to operate multiple RocksDB databases
> simultaneously. This is done via an environment object named Env. Among other
> things, a thread pool is associated with an Env. If applications want to
> share a common thread pool (for background compactions) among multiple
> database instances, then it should use the same Env object for opening those
> databases.
> Similarly, multiple database instances may share the same block cache.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
