[
https://issues.apache.org/jira/browse/FLINK-20496?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17317703#comment-17317703
]
Yu Li commented on FLINK-20496:
-------------------------------
[~maver1ck] Sorry that somehow I missed the ping and just noticed the message...
Normally we will only backport bug fix and operational improvements for minor
releases, and I think we could regard this one as the later case since it helps
to resolve perf-related (operational) problems. Would you like to open a new
JIRA and submit a PR against release-1.12 for the backport (since you've
already done this offline)? I will be there to help review and merge the
changes (smile).
And glad to know this work helps in your real-world workloads!
> RocksDB partitioned index filter option
> ---------------------------------------
>
> Key: FLINK-20496
> URL: https://issues.apache.org/jira/browse/FLINK-20496
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / State Backends
> Affects Versions: 1.10.2, 1.11.2, 1.12.0
> Reporter: YufeiLiu
> Assignee: YufeiLiu
> Priority: Major
> Labels: pull-request-available
> Fix For: 1.13.0
>
>
> When using RocksDBStateBackend and enabling
> {{state.backend.rocksdb.memory.managed}} and
> {{state.backend.rocksdb.memory.fixed-per-slot}}, flink will strictly limited
> rocksdb memory usage which contains "write buffer" and "block cache". With
> these options rocksdb stores index and filters in block cache, because in
> default options index/filters can grows unlimited.
> But it's lead another issue, if high-priority cache(configure by
> {{state.backend.rocksdb.memory.high-prio-pool-ratio}}) can't fit all
> index/filters blocks, it will load all metadata from disk when cache missed,
> and program went extremely slow. According to [Partitioned Index
> Filters|https://github.com/facebook/rocksdb/wiki/Partitioned-Index-Filters][1],
> we can enable two-level index having acceptable performance when
> index/filters cache missed.
> Enable these options can get over 10x faster in my case[2], I think we can
> add an option {{state.backend.rocksdb.partitioned-index-filters}} and default
> value is false, so we can use this feature easily.
> [1] Partitioned Index Filters:
> https://github.com/facebook/rocksdb/wiki/Partitioned-Index-Filters
> [2] Deduplicate scenario, state.backend.rocksdb.memory.fixed-per-slot=256M,
> SSD, elapsed time 4.91ms -> 0.33ms.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
