hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sean Sechrist (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HBASE-3046) EOFE replaying recovered.edits
Date Wed, 09 Feb 2011 22:07:57 GMT

    [ https://issues.apache.org/jira/browse/HBASE-3046?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12992738#comment-12992738
] 

Sean Sechrist commented on HBASE-3046:
--------------------------------------

We also saw this (hbase-0.89.20100924). It occurred shortly after the switches in our datacenter
restarted, so no nodes could communicate with each other. Our regionservers then started restarting
when they couldn't communicate with zookeeper. On regionserver restart, we saw that EOFException.
Here are the IRC logs from a chat with Stack about it: http://ssechrist.pastebin.com/PQ8YkpnY.
Relevant logs can be found in there. If anything else can help I'll post it.

> EOFE replaying recovered.edits
> ------------------------------
>
>                 Key: HBASE-3046
>                 URL: https://issues.apache.org/jira/browse/HBASE-3046
>             Project: HBase
>          Issue Type: Bug
>            Reporter: stack
>
> We're seeing an EOFE playing recovered.edits file loading region.  See http://permalink.gmane.org/gmane.comp.java.hadoop.hbase.user/12312
thread.  I took a look.  It looks like last record in file is incomplete as though we're not
flushing.

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message