You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@flink.apache.org by "Ted Yu (JIRA)" <ji...@apache.org> on 2018/10/02 01:57:00 UTC

[jira] [Comment Edited] (FLINK-7588) Document RocksDB tuning for spinning disks

    [ https://issues.apache.org/jira/browse/FLINK-7588?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16258309#comment-16258309 ] 

Ted Yu edited comment on FLINK-7588 at 10/2/18 1:56 AM:
--------------------------------------------------------

bq. Be careful about whether you have enough memory to keep all bloom filters

Other than the above being tricky, the other guidelines are actionable .


was (Author: yuzhihong@gmail.com):
bq. Be careful about whether you have enough memory to keep all bloom filters

Other than the above being tricky, the other guidelines are actionable.

> Document RocksDB tuning for spinning disks
> ------------------------------------------
>
>                 Key: FLINK-7588
>                 URL: https://issues.apache.org/jira/browse/FLINK-7588
>             Project: Flink
>          Issue Type: Improvement
>          Components: Documentation
>            Reporter: Ted Yu
>            Priority: Major
>              Labels: performance
>
> In docs/ops/state/large_state_tuning.md , it was mentioned that:
> bq. the default configuration is tailored towards SSDs and performs suboptimal on spinning disks
> We should add recommendation targeting spinning disks:
> https://github.com/facebook/rocksdb/wiki/RocksDB-Tuning-Guide#difference-of-spinning-disk



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)