[
https://issues.apache.org/jira/browse/HBASE-3102?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12920224#action_12920224
]
Gary Helmling commented on HBASE-3102:
--------------------------------------
For terminology, "marathon" seems a little opaque. How about "extendedperiod"
or something else that matches the current terminology?
Also seems a little odd that this applies only to RegionServerMetrics. Would
be really nice if this could extend the current Hadoop metrics functionality
and generically override refresh intervals per-metric. So you could do
something like: "hbase.period.compaction=3600"
Just an idea, but is there some way to extend
org.apache.hadoop.metrics.MetricsContext or org.apache.hadoop.metrics.Updater
to provide these kinds of overrides?
> Enhance HRegionServerMetrics for Long-running Stats
> ---------------------------------------------------
>
> Key: HBASE-3102
> URL: https://issues.apache.org/jira/browse/HBASE-3102
> Project: HBase
> Issue Type: Improvement
> Components: regionserver
> Reporter: Nicolas Spiegelberg
> Assignee: Nicolas Spiegelberg
> Fix For: 0.90.0
>
> Attachments: HBASE-3102_90.patch
>
>
> A useful metric to observe track is compaction count + duration. Since
> compactions are long running and only happen infrequently, the avg/opcount
> stats are should be reset longer than the default polling period (5 sec). In
> addition to 'hbase.period', we should allow a different duration after which
> long-running metrics should expire. This would also fix our existing metrics
> problem where min/max stats are never reset until the process is
> restarted/upgraded.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
