[
https://issues.apache.org/jira/browse/HBASE-12961?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14303826#comment-14303826
]
stack commented on HBASE-12961:
-------------------------------
Patch LGTM. Only issue is where to apply it. ServerLoad is public and we are
changing the return type so we'd 'break' compatibility. ServerLoad makes it
over to the client in ClusterStatus. This is a bug and ClusterStatus is more
admin/ops API. Where should we commit it? master branch for sure. branch-1 for
1.1.? We going to backport to 0.98?
> Negative values in read and write region server metrics
> --------------------------------------------------------
>
> Key: HBASE-12961
> URL: https://issues.apache.org/jira/browse/HBASE-12961
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Affects Versions: 2.0.0
> Reporter: Victoria
> Priority: Minor
> Attachments: HBASE-12961-2.0.0-v1.patch
>
>
> HMaster web page ui, shows the read/write request per region server. They are
> currently displayed by using 32 bit integers. Hence, if the servers are up
> for a long time the values can be shown as negative.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
