[
https://issues.apache.org/jira/browse/HBASE-16410?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ted Yu updated HBASE-16410:
---------------------------
Resolution: Won't Fix
Status: Resolved (was: Patch Available)
The cost of retrying distcp may be high (considering the amount of data to be
copied).
> Retry distcp job when previous distcp job didn't succeed
> --------------------------------------------------------
>
> Key: HBASE-16410
> URL: https://issues.apache.org/jira/browse/HBASE-16410
> Project: HBase
> Issue Type: Improvement
> Reporter: Ted Yu
> Assignee: Ted Yu
> Labels: backup
> Attachments: 16410.v1.txt
>
>
> Currently when copyService.copy() returns non-zero return value, we throw
> exception:
> {code}
> throw new IOException("Failed of Hadoop Distributed Copy from "+
> StringUtils.join(incrBackupFileList, ",") +" to "
> + backupContext.getHLogTargetDir());
> {code}
> We should retry distcp job when the return value is non-zero instead of
> throwing IOException immediately.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
