hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ted Yu (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-4742) Split dead server's log in parallel
Date Sun, 06 Nov 2011 02:44:51 GMT

    [ https://issues.apache.org/jira/browse/HBASE-4742?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13144895#comment-13144895
] 

Ted Yu commented on HBASE-4742:
-------------------------------

I don't see ProcessServerShutdown.java in 0.90 or TRUNK. So I assume it is from 0.89 branch.
I am not familiar with 0.89 branch. So my discussion below is tailored toward 0.92/TRUNK.

>From Liyin:
bq.  instead of distribute log splitting for each dead server as a separate thread.
I agree with the above statement.

In TRUNK, the method Liyin mentioned:
{code}
  public void splitLog(final List<ServerName> serverNames) throws IOException {
{code}
translates server names to log paths and uses the following to split the paths:
{code}
        splitLogSize = splitLogManager.splitLogDistributed(logDirs);
{code}
We can introduce a short interval for master to group the malfunctioning region servers so
that the above splitLog() API can be used.
If region servers go down with non-trivial interval in between, it is not obvious how the
above splitLog() API can be used.
                
> Split dead server's log in parallel
> -----------------------------------
>
>                 Key: HBASE-4742
>                 URL: https://issues.apache.org/jira/browse/HBASE-4742
>             Project: HBase
>          Issue Type: Improvement
>            Reporter: Liyin Tang
>            Assignee: Liyin Tang
>         Attachments: D237.1.patch, D237.2.patch, D237.3.patch, D237.4.patch
>
>
> When one region server goes down, the master will shutdown the region server and split
its log.
> However, splitting log is a blocking call and it would take some time.
> If more than one region server go down, the master will split its log one by one, which
is not efficient.
> Since we have the distributed log split, we could split these logs from the dead servers
in parallel. 

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message