You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@phoenix.apache.org by "Chao Wang (Jira)" <ji...@apache.org> on 2020/06/22 06:20:00 UTC
[jira] [Comment Edited] (PHOENIX-5971) Triple performance reduction
of delete operation from 4.4.0 Upgrade to 4.13.1
[ https://issues.apache.org/jira/browse/PHOENIX-5971?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17141656#comment-17141656 ]
Chao Wang edited comment on PHOENIX-5971 at 6/22/20, 6:19 AM:
--------------------------------------------------------------
Pass performance test， and analysis configuration. find a configuration influence. phoenix.mutate.batchSize.
this configuration default value from 1000 to 100 in 4.4.0 to 4.13.1. batch size smaller, batch more and more, hbase rpc increase. and then block rpc request. so triple performance reduction.
I think we should set phoenix.mutate.batchSize in hbase-site.xml according to use scenario. if you need large writing operation, you set phoenix.mutate.batchSize large.
was (Author: wangchao316):
Pass performance test， and analysis configuration. find a configuration influence. phoenix.mutate.batchSize.
this configuration from 1000 to 100 in 4.13.1. batch size smaller, batch more and more, hbase rpc increase. and then block rpc request. so triple performance reduction.
I think we should set phoenix.mutate.batchSize in hbase-site.xml according to use scenario. if you need large writing operation, you set phoenix.mutate.batchSize large.
> Triple performance reduction of delete operation from 4.4.0 Upgrade to 4.13.1
> -----------------------------------------------------------------------------
>
> Key: PHOENIX-5971
> URL: https://issues.apache.org/jira/browse/PHOENIX-5971
> Project: Phoenix
> Issue Type: Bug
> Components: core
> Affects Versions: 4.13.1
> Environment: 200+ node
> Reporter: Chao Wang
> Assignee: Chao Wang
> Priority: Major
>
> phoenix from 4.4.0 Upgrade to 4.13.1 in my environmental, delete operation triple performance reduction. for example a sql (delete ... from ...) will delete million of data, and then hundreds of concurrent.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)