[jira] [Commented] (HBASE-19344) improve asyncWAL by using Independent thread for netty #IO in FanOutOneBlockAsyncDFSOutput

["Chance Li (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HBASE-19344?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16265684#comment-16265684 ] 

Chance Li commented on HBASE-19344:
-----------------------------------

Some of the related unit tests have passed now. There are still total unit tests and ITBLL need to be finished.
[~Apache9] pls help to review
the result:
wal=1, 1 col 1024B, 30 regions	ycsb: 4*10*100	
FSHLog(114157) -> AsynWal No Patch(99062) -> AsynWal With Patch(114178)
!wal-1-test-result.png!

 wal=8, 1 col 1024B, 30 regions	ycsb: 4*10*100 .
FSHLog(138053) -> AsynWal No Patch(TBD) -> AsynWal With Patch, only 1 thread #IO(38426) -> AsynWal With Patch, 16 thread #IO(192707)
!wal-8-test-result.png!
config:
hbase.wal.regiongrouping.delegate.provider=asyncfs or defaultProvider
hbase.wal.provider=multiwal
hbase.wal.regiongrouping.numgroups=8 or 1
hbase.wal.async.netty.worker.count=1 or 16

> improve asyncWAL by using Independent thread for netty #IO in FanOutOneBlockAsyncDFSOutput 
> -------------------------------------------------------------------------------------------
>
>                 Key: HBASE-19344
>                 URL: https://issues.apache.org/jira/browse/HBASE-19344
>             Project: HBase
>          Issue Type: Improvement
>          Components: wal
>    Affects Versions: 2.0.0-beta-1
>            Reporter: Chance Li
>            Assignee: Chance Li
>             Fix For: 2.0.0
>
>         Attachments: wal-1-test-result.png, wal-8-test-result.png
>
>
> The logic now is that the netty #IO thread and asyncWal's thread are the same one.
> Improvement proposal:
> 1, Split into two.
> 2, All multiWal share the netty #IO thread pool. 



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)