You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Jean-Daniel Cryans (JIRA)" <ji...@apache.org> on 2010/11/02 01:22:25 UTC
[jira] Updated: (HBASE-3179) Enable ReplicationLogsCleaner only if
replication is, and fix its test
[ https://issues.apache.org/jira/browse/HBASE-3179?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jean-Daniel Cryans updated HBASE-3179:
--------------------------------------
Issue Type: Bug (was: Improvement)
Summary: Enable ReplicationLogsCleaner only if replication is, and fix its test (was: Once a minute, master logs being filled with ReplicationLogsCleaner messages even if no replication in use)
> Enable ReplicationLogsCleaner only if replication is, and fix its test
> ----------------------------------------------------------------------
>
> Key: HBASE-3179
> URL: https://issues.apache.org/jira/browse/HBASE-3179
> Project: HBase
> Issue Type: Bug
> Components: master, replication
> Affects Versions: 0.90.0
> Reporter: Jonathan Gray
> Assignee: Jean-Daniel Cryans
> Priority: Critical
> Fix For: 0.90.0
>
>
> I'm seeing logs like this on a trunk master:
> {noformat}
> 2010-10-31 11:21:13,793 DEBUG org.apache.hadoop.hbase.replication.master.ReplicationLogCleaner: Didn't find this log in ZK, deleting: 1.1.1.1%3A60020.1288545516340
> 2010-10-31 11:21:13,793 WARN org.apache.hadoop.hdfs.DFSClient: File /HBASE/.oldlogs/1.1.1.1%3A60020.1288545516340 is beng deleted only through Trash org.apache.hadoop.fs.FsShell.delete because all deletes must go through Trash.
> 2010-10-31 11:21:13,798 DEBUG org.apache.hadoop.hbase.replication.master.ReplicationLogCleaner: Didn't find this log in ZK, deleting: 1.1.1.1%3A60020.1288545899208
> 2010-10-31 11:21:13,798 WARN org.apache.hadoop.hdfs.DFSClient: File /HBASE/.oldlogs/1.1.1.1%3A60020.1288545899208 is beng deleted only through Trash org.apache.hadoop.fs.FsShell.delete because all deletes must go through Trash.
> {noformat}
> There are like 50 of these showing up once a minute. I'm unsure if this is more valuable when replication is turned on, but should this be running when replication is not on in this cluster? At the least there should be less logging, like one line per run.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.