[
https://issues.apache.org/jira/browse/HBASE-16009?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15325420#comment-15325420
]
Ted Yu commented on HBASE-16009:
--------------------------------
In my experiment, backup_1465575766499 was an incremental backup.
I first restored full backup (backup rootdir was different).
Restoring incremental backup should not require -overwrite.
Otherwise user cannot obtain the data from both backup's.
> Restoring an incremental backup should not require -overwrite
> -------------------------------------------------------------
>
> Key: HBASE-16009
> URL: https://issues.apache.org/jira/browse/HBASE-16009
> Project: HBase
> Issue Type: Bug
> Reporter: Ted Yu
> Labels: backup
> Attachments: 16009.v1.txt
>
>
> When I tried to restore an incremental backup,
> hbase restore hdfs://hbase-test-rc-rerun-6:8020/user/hbase
> backup_1465575766499 t1 t2
> I got:
> {code}
> 2016-06-10 19:53:11,317 ERROR [main] impl.RestoreClientImpl: ERROR: restore
> failed with error: Existing table found in target while no "-overwrite"
> option found
> java.io.IOException: Existing table found in target while no "-overwrite"
> option found
> at
> org.apache.hadoop.hbase.backup.impl.RestoreClientImpl.checkTargetTables(RestoreClientImpl.java:186)
> at
> org.apache.hadoop.hbase.backup.impl.RestoreClientImpl.restore(RestoreClientImpl.java:108)
> at
> org.apache.hadoop.hbase.backup.RestoreDriver.parseAndRun(RestoreDriver.java:169)
> at
> org.apache.hadoop.hbase.backup.RestoreDriver.doWork(RestoreDriver.java:198)
> {code}
> The above check should only be performed for restoring full backup.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
