[
https://issues.apache.org/jira/browse/HDFS-1969?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Todd Lipcon updated HDFS-1969:
------------------------------
Attachment: hdfs-1969.txt
Here's a patch for trunk which fixes the issue (verified with unit test and
manually).
There were a couple problems with the test:
- it didn't check the exception text at all, so some of the cases were actually
failing due to an unrelated issue (the "previous" dir ended up with md5
checksums even though the version was prior to that feature)
- one of the test cases ran the NN in upgrade mode instead of rollback mode, so
wasn't actually testing what was intended.
> Running rollback on new-version namenode destroys namespace
> -----------------------------------------------------------
>
> Key: HDFS-1969
> URL: https://issues.apache.org/jira/browse/HDFS-1969
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: name-node
> Affects Versions: 0.22.0
> Reporter: Todd Lipcon
> Assignee: Todd Lipcon
> Priority: Blocker
> Fix For: 0.22.0
>
> Attachments: hdfs-1969.txt
>
>
> The following sequence leaves the namespace in an inconsistent/broken state:
> - format NN using 0.20 (or any prior release, probably)
> - run hdfs namenode -upgrade on 0.22. ^C the NN once it comes up.
> - run hdfs namenode -rollback on 0.22 (this should fail but doesn't!)
> This leaves the name directory in a state such that the version file claims
> it's an 0.20 namespace, but the fsimage is in 0.22 format. It then crashes
> when trying to start up.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
