[
https://issues.apache.org/jira/browse/HDFS-12420?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16168160#comment-16168160
]
Daryn Sharp commented on HDFS-12420:
------------------------------------
I'll preface with I think this is inane. This is like linux having a sysctl
controlling whether you can use rm -r.
That said, if we really are going to make this change:
* Clarify in the interactive message that all data will be destroyed. I see
the patch does that and I think it should be sufficient. But...
* If there's a conf, "hadoop.is.prod.cluster" is silly. Call it something more
like "hdfs.reformat.enabled".
* Instead of a conf, maybe consider the presence of some file in the metadata
directory as the protection.
* I'd really rather allow -force work otherwise it's not really a force. If
you run a "format" command with a "-force" then you made a huge conscious
decision to wipe your NN. You can't typo that.
> Disable Namenode format for prod clusters when data already exists
> ------------------------------------------------------------------
>
> Key: HDFS-12420
> URL: https://issues.apache.org/jira/browse/HDFS-12420
> Project: Hadoop HDFS
> Issue Type: Improvement
> Reporter: Ajay Kumar
> Assignee: Ajay Kumar
> Attachments: HDFS-12420.01.patch, HDFS-12420.02.patch,
> HDFS-12420.03.patch, HDFS-12420.04.patch, HDFS-12420.05.patch
>
>
> Disable NameNode format to avoid accidental formatting of Namenode in
> production cluster. If someone really wants to delete the complete fsImage,
> they can first delete the metadata dir and then run {code} hdfs namenode
> -format{code} manually.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
