hbase-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Andrew Purtell (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HBASE-1982) [EC2] handle potentially large and uneven instance startup times
Date Fri, 18 Dec 2009 23:33:18 GMT

    [ https://issues.apache.org/jira/browse/HBASE-1982?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12792695#action_12792695
] 

Andrew Purtell commented on HBASE-1982:
---------------------------------------

The simplest and most direct approach is to support a new command line parameter for the master
which sets the minimum number of region servers that must check in before user tables are
deployed. 

> [EC2] handle potentially large and uneven instance startup times
> ----------------------------------------------------------------
>
>                 Key: HBASE-1982
>                 URL: https://issues.apache.org/jira/browse/HBASE-1982
>             Project: Hadoop HBase
>          Issue Type: Sub-task
>            Reporter: Andrew Purtell
>            Assignee: Andrew Purtell
>            Priority: Minor
>             Fix For: 0.20.3, 0.21.0
>
>
> The variability in instance start latency may impact a HBase cluster. All the slaves
are started with one invocation of ec2-run-instances but not all may come up at once. Handle
potentially large and uneven instance startup times through a master configuration option
that prevents assignment of user tables until the configured number of region servers are
up. 

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


Mime
View raw message