[
https://issues.apache.org/jira/browse/HBASE-13516?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14516483#comment-14516483
]
Hudson commented on HBASE-13516:
--------------------------------
FAILURE: Integrated in HBase-1.2 #33 (See
[https://builds.apache.org/job/HBase-1.2/33/])
HBASE-13516 Increase PermSize to 128MB (enis: rev
cf3e5ebf599a2a4c8ce9e18608c3947bbdce6f5f)
* conf/hbase-env.cmd
* conf/hbase-env.sh
> Increase PermSize to 128MB
> --------------------------
>
> Key: HBASE-13516
> URL: https://issues.apache.org/jira/browse/HBASE-13516
> Project: HBase
> Issue Type: Improvement
> Reporter: Enis Soztutar
> Assignee: Enis Soztutar
> Fix For: 2.0.0, 1.1.0, 1.2.0
>
> Attachments: hbase-13516_v1.patch, hbase-13516_v2.patch
>
>
> HBase uses ~40MB, and with Phoenix we use ~56MB of Perm space out of 64MB by
> default. Every Filter and Coprocessor increases that.
> Running out of perm space triggers a stop the world full GC of the entire
> heap. We have seen this in misconfigured cluster.
> Should we default to {{-XX:PermSize=128m -XX:MaxPermSize=128m}} out of the
> box as a convenience for users?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
