accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Keith Turner (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (ACCUMULO-3480) GroupBalancer improvements
Date Thu, 22 Jan 2015 00:38:35 GMT

    [ https://issues.apache.org/jira/browse/ACCUMULO-3480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14286655#comment-14286655
] 

Keith Turner commented on ACCUMULO-3480:
----------------------------------------


Do you think we need to try and optimize this?

I modified the {{bigTest()}} unit test in the patch to have 1,000 groups (so 241,000 tablets).
  Then I timed balance(5000) twice.  The 1st time it took 5.1 secs (there were 34,304 migrations).
 The 2nd time (when everything was balanced and 0 migrations happened) it took .95 secs. 
  So in this particular case the balancer would probably chew up at least 1 sec of cpu time
every 60 secs (assuming thats how often the balancer is configured to run).  I think thats
ok.

> GroupBalancer improvements
> --------------------------
>
>                 Key: ACCUMULO-3480
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-3480
>             Project: Accumulo
>          Issue Type: Bug
>            Reporter: Keith Turner
>            Assignee: Keith Turner
>             Fix For: 1.7.0
>
>         Attachments: ACCUMULO-3480-1.patch
>
>
> [~bfloss] made the following comment on ACCUMULO-3439
> bq. The balance method in GroupBalancer does nothing if there are any incoming migrations.
You probably want to check the incoming migrations and only bail out if any of those migrations
are for the specific table that the GroupBalancer is managing.
> I chatted w/ Brian and he also mentioned that the group balancer should record time at
the end, and not the beginning.  



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

Mime
View raw message