[
https://issues.apache.org/jira/browse/KUDU-2726?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17248727#comment-17248727
]
wangningito commented on KUDU-2726:
-----------------------------------
Ah, I see, thx for point.
The problem would be much clearer in this route.
I'm comping up with this: having a slide window with fixed size(maybe can be
adjusted with runtime flag), slide with height calculation to find the local
optima. And then expose some related metrics to help user to do adjustment on
this runtime flag.
And to a compaction, it's final target is to solve the problem of 'rowset
overlapping', if any subset has local optima is reached, then local optimal can
also be see as global optima.
> Very large tablets defeat budgeted compaction
> ---------------------------------------------
>
> Key: KUDU-2726
> URL: https://issues.apache.org/jira/browse/KUDU-2726
> Project: Kudu
> Issue Type: Improvement
> Affects Versions: 1.9.0
> Reporter: William Berkeley
> Priority: Major
> Labels: density, roadmap-candidate
>
> On very large tablets (50GB+), despite being very uncompacted with a large
> average rowset height, a default budget (128MB) worth of compaction may not
> reduce average rowset height enough to pass the minimum threshold. Thus the
> tablet stays uncompacted forever.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
