ambari-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Nick Dimiduk (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (AMBARI-11290) Avg Load for HBase is 261 regions per RS for a fresh one node cluster
Date Fri, 22 May 2015 18:30:17 GMT

    [ https://issues.apache.org/jira/browse/AMBARI-11290?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14556564#comment-14556564
] 

Nick Dimiduk commented on AMBARI-11290:
---------------------------------------

>From HBase perspective, 250 regions on a single RS should be presented by Ambari as at
least a warning, if not critical state. It's definitely the upper limit in terms of what HBase
community recommends. Maybe some folks like to run their clusters "hot", so they might want
to configure the alert thresholds.

>From Phoenix perspective, 250 regions for this table, all hosted on a small number (1-3)
or hosts is actually going to be a performance penalty; opposite of the intended effect.

> Avg Load for HBase is 261 regions per RS for a fresh one node cluster
> ---------------------------------------------------------------------
>
>                 Key: AMBARI-11290
>                 URL: https://issues.apache.org/jira/browse/AMBARI-11290
>             Project: Ambari
>          Issue Type: Task
>          Components: ambari-server
>    Affects Versions: 2.1.0
>            Reporter: Siddharth Wagle
>            Assignee: Siddharth Wagle
>             Fix For: 2.1.0
>
>
> This happens because newer version of Phoenix sets salBuckets for sequence table to 256
creating 256 regions for salting by default.
> We can add the above config to Ambari's computed values – anything smaller than 10
RS's should use a value of $numRs * 5. Something like this. It could be a hassle for folks
who start small and add nodes; they'll need to split the table.



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

Mime
View raw message