[ 
https://issues.apache.org/jira/browse/FLINK-8639?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16360822#comment-16360822
 ] 

ASF GitHub Bot commented on FLINK-8639:
---------------------------------------

GitHub user sihuazhou opened a pull request:

    https://github.com/apache/flink/pull/5465

    [FLINK-8639][State Backends]Fix always need to seek multiple times when 
iterator RocksDBMapState

    ## What is the purpose of the change
    
    This PR addressed issue 
[FLINK-8639](https://issues.apache.org/jira/browse/FLINK-8639). Currently, 
almost every time we want to iterator a RocksDBMapState we need to do seek at 
least 2 times (Seek is a poor performance action for rocksdb cause it can't use 
the bloomfilter). This is because `RocksDBMapIterator` use a `cacheEntries` to 
cache the seek values every time and the `cacheEntries`'s init size is 1.
    
    ## Brief change log
    
    
    - Change CACHE_SIZE_BASE from 1 to 32.
    - Compare in reverse order in function underSameKey().
    
    ## Verifying this change
    
    This change is already covered by existing tests.
    
    ## Does this pull request potentially affect one of the following parts:
    
      - Dependencies (does it add or upgrade a dependency): (no)
      - The public API, i.e., is any changed class annotated with 
`@Public(Evolving)`: (no)
      - The serializers: (no)
      - The runtime per-record code paths (performance sensitive): (no)
      - Anything that affects deployment or recovery: JobManager (and its 
components), Checkpointing, Yarn/Mesos, ZooKeeper: (no)
      - The S3 file system connector: (no)
    
    ## Documentation
    
      - Does this pull request introduce a new feature? (no)


You can merge this pull request into a Git repository by running:

    $ git pull https://github.com/sihuazhou/flink improve_rocksdb_base_mapstate

Alternatively you can review and apply these changes as the patch at:

    https://github.com/apache/flink/pull/5465.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

    This closes #5465
    
----
commit 753855ec2957492ba6c2eb272fdc5ffe7c769608
Author: summerleafs <summerleafs@...>
Date:   2018-02-12T14:28:04Z

    1, Removed CACHE_SIZE_BASE, CACHE_SIZE_LIMIT.
    2, Compare in reverse order in function underSameKey().

commit e33b807ba151265e0fd86d52701965dc5ffcb7be
Author: summerleafs <summerleafs@...>
Date:   2018-02-12T14:50:28Z

    change CACHE_SIZE_BASE from 1 to 32.

----


> Fix always need to seek multiple times when iterator RocksDBMapState
> --------------------------------------------------------------------
>
>                 Key: FLINK-8639
>                 URL: https://issues.apache.org/jira/browse/FLINK-8639
>             Project: Flink
>          Issue Type: Improvement
>          Components: State Backends, Checkpointing
>    Affects Versions: 1.4.0
>            Reporter: Sihua Zhou
>            Assignee: Sihua Zhou
>            Priority: Critical
>             Fix For: 1.5.0
>
>
> Currently, almost every time we want to iterator a RocksDBMapState we need to 
> do seek at least 2 times (Seek is a poor performance action for rocksdb cause 
> it can't use the bloomfilter). This is because `RocksDBMapIterator` use a 
> `cacheEntries` to cache the seek values every time and the `cacheEntries`'s 
> init size is 1.



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

Reply via email to