ambari-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sumit Mohanty (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (AMBARI-18574) Set hbase.hregion.memstore.chunkpool.maxsize to 1.0
Date Thu, 24 Aug 2017 14:20:01 GMT

    [ https://issues.apache.org/jira/browse/AMBARI-18574?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16140071#comment-16140071
] 

Sumit Mohanty commented on AMBARI-18574:
----------------------------------------

The patch, as such, looks good. Not sure if its intentional but with this change when Ambari
is upgraded the property will show up in HBase config and require HBase restart. In general,
HBase restart is not as problematic as HDFS but most users do not like to restart any service
after Ambari upgrade. So if you want to avoid HBase restart after Ambari upgrade and rather
"fix" this as part of a stack upgrade then you should set {{<on-ambari-upgrade add="false"/>}}
and make upgrade pack changes to add the config when stack is upgraded.

> Set hbase.hregion.memstore.chunkpool.maxsize to 1.0
> ---------------------------------------------------
>
>                 Key: AMBARI-18574
>                 URL: https://issues.apache.org/jira/browse/AMBARI-18574
>             Project: Ambari
>          Issue Type: Bug
>            Reporter: Ted Yu
>            Assignee: Yao Lei
>             Fix For: 2.6.0
>
>         Attachments: AMBARI-18574.patch
>
>
> The default value for hbase.hregion.memstore.chunkpool.maxsize is:
> {code}
>   final static float POOL_MAX_SIZE_DEFAULT = 0.0f;
> {code}
> This would result in chunk pool being disabled, leading to excessive MemStoreLAB chunk
allocations.
> Ambari should set the value to 1.0



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

Mime
View raw message