[
https://issues.apache.org/jira/browse/HBASE-15055?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15091565#comment-15091565
]
Anoop Sam John commented on HBASE-15055:
----------------------------------------
The main concern was that because of this size limit, even major compaction is
skipping these large files. (ie. irrespective of TTL is there or not I
believe)..
The fix avoids the size check based on TTL and expired ratio. I agree that
later on there are still some more logic to exclude files. Still things look
too much complicated now... My Qs
1. When this max size is specified, at least the major compaction should
select all files and do a real major compaction right? Else when those will
get compacted? Possible tomb stone removal! This should happen irrespective
of the TTL thing no?
2. When this size limit is specified, that should be respected for at least
minor compaction. Here there is chance that it can get violated.. (With even
not a single cell TTL expired within that big file)
Still I feel some thing problematic in the whole.. We might need better
refactoring of these things rather than adding new configs again and again.
> Major compaction is not triggered when both of TTL and
> hbase.hstore.compaction.max.size are set
> -----------------------------------------------------------------------------------------------
>
> Key: HBASE-15055
> URL: https://issues.apache.org/jira/browse/HBASE-15055
> Project: HBase
> Issue Type: Bug
> Reporter: Eungsop Yoo
> Assignee: Eungsop Yoo
> Priority: Minor
> Attachments: HBASE-15055-v1.patch, HBASE-15055-v2.patch,
> HBASE-15055-v3.patch, HBASE-15055-v4.patch, HBASE-15055-v5.patch,
> HBASE-15055-v6.patch, HBASE-15055-v7.patch, HBASE-15055-v8.patch,
> HBASE-15055-v9.patch, HBASE-15055.patch
>
>
> Some large files may be skipped by hbase.hstore.compaction.max.size in
> candidate selection. It causes skipping of major compaction. So the TTL
> expired records are still remained in the disks and keep consuming disks.
> To resolve this issue, I suggest that to skip large files only if there is no
> TTL expired record.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
