[jira] [Updated] (HBASE-17049) Find out why AsyncFSWAL issues much more syncs than FSHLog

["Duo Zhang (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HBASE-17049?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Duo Zhang updated HBASE-17049:
------------------------------
    Attachment: delay-sync.patch

Tried to delay sync so we can aggregate more appends before actually issue a sync. It could reduce the sync count greatly in WALPE with a 1ms delay, but the performance is also reduced... And for PE, I can not get a better result either...

I also set the initial buffer size to 80KB in FanOutOneBlockDFSOutput as there is no progress in HBASE-17048.

[~stack] [~ram_krish] You guys could have a try to see if it helps. And you guys could also try changing the batchSize to see if it helps. For me it does not help either but [~ram_krish] said use a large batchSize could get a better result for him.

Thanks.

> Find out why AsyncFSWAL issues much more syncs than FSHLog
> ----------------------------------------------------------
>
>                 Key: HBASE-17049
>                 URL: https://issues.apache.org/jira/browse/HBASE-17049
>             Project: HBase
>          Issue Type: Sub-task
>          Components: wal
>    Affects Versions: 2.0.0
>            Reporter: Duo Zhang
>             Fix For: 2.0.0
>
>         Attachments: delay-sync.patch
>
>
> https://issues.apache.org/jira/browse/HBASE-16890?focusedCommentId=15647590&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-15647590



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)