[jira] [Commented] (HBASE-14460) [Perf Regression] Merge of MVCC and SequenceId (HBASE-HBASE-8763) slowed Increments, CheckAndPuts, batch operations

["Elliott Clark (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HBASE-14460?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14902002#comment-14902002 ] 

Elliott Clark commented on HBASE-14460:
---------------------------------------

Yeah this is mostly because we have no idea what rows we're waiting for on mvcc. So to be extra cautious we wait for everything.
After HBASE-12751 goes in we can hold the row lock until after sync (since it's reader/writer this shouldn't slow down normal puts). That would mean that getting the write rowlock means that there are no transactions for that row in flight. That would make increment/check and mutate a little bit faster.

> [Perf Regression] Merge of MVCC and SequenceId (HBASE-HBASE-8763) slowed Increments, CheckAndPuts, batch operations
> -------------------------------------------------------------------------------------------------------------------
>
>                 Key: HBASE-14460
>                 URL: https://issues.apache.org/jira/browse/HBASE-14460
>             Project: HBase
>          Issue Type: Bug
>          Components: Performance
>            Reporter: stack
>            Assignee: stack
>            Priority: Critical
>         Attachments: 14460.txt, region_lock.png
>
>
> As reported by 鈴木俊裕 up on the mailing list -- see "Performance degradation between CDH5.3.1(HBase0.98.6) and CDH5.4.5(HBase1.0.0)" -- our unification of sequenceid and MVCC slows Increments (and other ops) as the mvcc needs to 'catch up' to our current point before we can read the last Increment value that we need to update.
> We can say that our Increment is just done wrong, we should just be writing Increments and summing on read, but checkAndPut as well as batching operations have the same issue. Fix.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)