tephra-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Lars Hofhansl (JIRA)" <j...@apache.org>
Subject [jira] [Comment Edited] (TEPHRA-299) Executing a large batch delete is very slow
Date Fri, 24 May 2019 21:10:00 GMT

    [ https://issues.apache.org/jira/browse/TEPHRA-299?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16847884#comment-16847884
] 

Lars Hofhansl edited comment on TEPHRA-299 at 5/24/19 9:09 PM:
---------------------------------------------------------------

-complete does the same for hbase 1.3, 1.4, and 2.0. I think we do not have to go back and
make this change for any earlier versions (it's a perf fix and few folks will use the old
versions anyway).
 Please have a quick looksee (it's the same change, just also adapted to HBase 1.3 and 2.0.)

Unless I hear objections I'll try to commit this in a bit (I see I'm listed as a mentor, not
committer, so I might not have the commit mojo after all).


was (Author: lhofhansl):
-complete does the same for hbase 1.3 and 2.0. I think we do not have to go back and make
this change for any earlier versions (it's a perf fix and few folks will use the old versions
anyway).
Please have a quick looksee (it's the same change, just also adapted to HBase 1.3 and 2.0.)

Unless I hear objections I'll try to commit this in a bit (I see I'm listed as a mentor, not
committer, so I might not have the commit mojo after all).


> Executing a large batch delete is very slow
> -------------------------------------------
>
>                 Key: TEPHRA-299
>                 URL: https://issues.apache.org/jira/browse/TEPHRA-299
>             Project: Tephra
>          Issue Type: Bug
>    Affects Versions: 0.15.0-incubating
>            Reporter: Lars Hofhansl
>            Assignee: Lars Hofhansl
>            Priority: Major
>         Attachments: 299-DOES-NOT_WORK.txt, 299-client-v2.txt, 299-client-v3.txt, 299-client.txt,
299-complete.txt
>
>
> I noticed that batch deletes are quire slow. In the profiler I found that almost all
of the time is spent in org.apache.hadoop.hbase.regionserver.wal.FSHLog.blockOnSync().
> Looking at TransactionProcessor.preDelete it is obvious why:
> The batch delete is translated into *single* puts that are added to the region one by
one, so each time the WAL is flushed.
>  



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

Mime
View raw message