hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Duo Zhang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-16890) Analyze the performance of AsyncWAL and fix the same
Date Thu, 23 Mar 2017 02:40:41 GMT

    [ https://issues.apache.org/jira/browse/HBASE-16890?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15937594#comment-15937594
] 

Duo Zhang commented on HBASE-16890:
-----------------------------------

[~stack] [~ram_krish] Let's pick this up again?

The last consensus in HBASE-17049 is that, using multiwal can increase the performance of
AsyncFSWAL more than FSHLog. [~stack] Could you please also verify it sir? I think the problem
is that AsyncFSWAL is single threaded but FSHLog has 5 sync threads.

And for me, I will run PE tool against a distributed cluster next. Maybe a single RS but HDFS
will be distributed. I think this is the common scenario in the real world.

Thanks.

> Analyze the performance of AsyncWAL and fix the same
> ----------------------------------------------------
>
>                 Key: HBASE-16890
>                 URL: https://issues.apache.org/jira/browse/HBASE-16890
>             Project: HBase
>          Issue Type: Sub-task
>          Components: wal
>    Affects Versions: 2.0.0
>            Reporter: ramkrishna.s.vasudevan
>            Assignee: ramkrishna.s.vasudevan
>             Fix For: 2.0.0
>
>         Attachments: async.svg, AsyncWAL_disruptor_1 (2).patch, AsyncWAL_disruptor_3.patch,
AsyncWAL_disruptor_3.patch, AsyncWAL_disruptor_4.patch, AsyncWAL_disruptor_6.patch, AsyncWAL_disruptor.patch,
classic.svg, contention_defaultWAL.png, contention.png, HBASE-16890-rc-v2.patch, HBASE-16890-rc-v3.patch,
HBASE-16890-remove-contention.patch, HBASE-16890-remove-contention-v1.patch, Screen Shot 2016-10-25
at 7.34.47 PM.png, Screen Shot 2016-10-25 at 7.39.07 PM.png, Screen Shot 2016-10-25 at 7.39.48
PM.png, Screen Shot 2016-11-04 at 5.21.27 PM.png, Screen Shot 2016-11-04 at 5.30.18 PM.png
>
>
> Tests reveal that AsyncWAL under load in single node cluster performs slower than the
Default WAL. This task is to analyze and see if we could fix it.
> See some discussions in the tail of JIRA HBASE-15536.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message