You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hbase.apache.org by "stack (JIRA)" <ji...@apache.org> on 2008/03/13 08:56:41 UTC
[jira] Resolved: (HBASE-508) regionserver stuck log rolling and
doing nothing else
[ https://issues.apache.org/jira/browse/HBASE-508?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
stack resolved HBASE-508.
-------------------------
Resolution: Invalid
Server does not go on to OOME as we've seen in the past. Later in the log, the master deployed a new region to this server. This seemed to shake things up. After all of the opening hysteria around deploy of new region, we did a couple of flushes on an old region -- were 'normal' 64M in size -- then we went on to do a compaction and a cleanup of most of the logs we'd just written. It looks like nothing actually wrong here.
Closing as invalid. Will improve quality of logging as part of another issue.
> regionserver stuck log rolling and doing nothing else
> -----------------------------------------------------
>
> Key: HBASE-508
> URL: https://issues.apache.org/jira/browse/HBASE-508
> Project: Hadoop HBase
> Issue Type: Bug
> Reporter: stack
> Attachments: td.log
>
>
> When I look in the log of XX.XX.XX.92, it shows only log rolling activity of following form:
> {code}
> ...
> 2008-03-13 07:04:17,596 DEBUG org.apache.hadoop.hbase.HLog: Closing current log writer hdfs://coral-dfs.cluster.powerset.com:10000/hbase/XX.XX.XX-2.u.powerset.com/log_XX.XX.XX.92_1205384328364_60020/hlog.dat.022 to get a new one
> 2008-03-13 07:04:17,599 INFO org.apache.hadoop.hbase.HLog: new log writer created at hdfs://coral-dfs.cluster.powerset.com:10000/hbase/XX.XX.XX-2.u.powerset.com/log_XX.XX.XX.92_1205384328364_60020/hlog.dat.023
> 2008-03-13 07:08:48,425 INFO org.apache.hadoop.hbase.HRegionServer: Rolling hlog. Number of entries: 30004
> 2008-03-13 07:08:48,472 DEBUG org.apache.hadoop.hbase.HLog: Closing current log writer hdfs://coral-dfs.cluster.powerset.com:10000/hbase/XX.XX.XX-2.u.powerset.com/log_XX.XX.XX.92_1205384328364_60020/hlog.dat.023 to get a new one
> 2008-03-13 07:08:48,484 INFO org.apache.hadoop.hbase.HLog: new log writer created at hdfs://coral-dfs.cluster.powerset.com:10000/hbase/XX.XX.XX-2.u.powerset.com/log_XX.XX.XX.92_1205384328364_60020/hlog.dat.024
> ~
> ...
> {code}
> .. which is odd because if we're taking on this many edits, you'd think there'd be flushing and compacting going on. There's none.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.