[ 
https://issues.apache.org/jira/browse/HBASE-2353?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12877748#action_12877748
 ] 

HBase Review Board commented on HBASE-2353:
-------------------------------------------

Message from: "Todd Lipcon" <[email protected]>

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://review.hbase.org/r/167/
-----------------------------------------------------------

Review request for hbase, Kannan Muthukkaruppan and Ryan Rawson.


Summary
-------

I implemented the "mini batching" idea we talked about on the JIRA.

This currently breaks some of the error handling, so I dont intend to commit as 
is, but everyone is busy so wanted to put a review up now while I tidy up the 
rest.


This addresses bug HBASE-2353.
    http://issues.apache.org/jira/browse/HBASE-2353


Diffs
-----

  src/main/java/org/apache/hadoop/hbase/regionserver/HRegion.java 6b6d098 
  src/main/java/org/apache/hadoop/hbase/regionserver/HRegionServer.java a1baff4 
  src/main/java/org/apache/hadoop/hbase/regionserver/wal/HLog.java 034690e 

Diff: http://review.hbase.org/r/167/diff


Testing
-------

Some PEs on a real sync-enabled cluster, seems faster but haven't done 
scientific benchmarking.


Thanks,

Todd




> HBASE-2283 removed bulk sync optimization for multi-row puts
> ------------------------------------------------------------
>
>                 Key: HBASE-2353
>                 URL: https://issues.apache.org/jira/browse/HBASE-2353
>             Project: HBase
>          Issue Type: Bug
>            Reporter: ryan rawson
>            Assignee: Todd Lipcon
>            Priority: Blocker
>             Fix For: 0.21.0
>
>         Attachments: HBASE-2353_def_log_flush.patch
>
>
> previously to HBASE-2283 we used to call flush/sync once per put(Put[]) call 
> (ie: batch of commits).  Now we do for every row.  
> This makes bulk uploads slower if you are using WAL.  Is there an acceptable 
> solution to achieve both safety and performance by bulk-sync'ing puts?  Or 
> would this not work in face of atomic guarantees?
> discuss!

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to