[
https://issues.apache.org/jira/browse/HBASE-15205?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15138300#comment-15138300
]
Enis Soztutar commented on HBASE-15205:
---------------------------------------
bq. If we are ok with the suggestion of adding all the families that has non
default scope per WAL Key then we can make things much faster in the direct
write path since what ever we extract from the HRegion can directly be
persisted per WALKey. No need for any checks in the write path to see what the
families with each cell and then creating another local map that can be
persisted in PBs.
I was reading the patch without looking further down the comments and was
thinking the same thing. I think it is worth doing it. Although it is not just
the WAL overhead, but the replicate RPC also encodes the scopes.
> Do not find the replication scope for every WAL#append()
> --------------------------------------------------------
>
> Key: HBASE-15205
> URL: https://issues.apache.org/jira/browse/HBASE-15205
> Project: HBase
> Issue Type: Sub-task
> Components: regionserver
> Reporter: ramkrishna.s.vasudevan
> Assignee: ramkrishna.s.vasudevan
> Priority: Minor
> Fix For: 2.0.0
>
> Attachments: HBASE-15205.patch, ScopeWALEdits.jpg,
> ScopeWALEdits_afterpatch.jpg
>
>
> After the byte[] and char[] the other top contributor for lot of GC (though
> it is only 2.86%) is the UTF_8.newDecoder.
> This happens because for every WAL append we try to calculate the replication
> scope associate with the families associated with the TableDescriptor. I
> think per WAL append doing this is very costly and creates lot of garbage.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
