[ 
https://issues.apache.org/jira/browse/HDFS-1111?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12861153#action_12861153
 ] 

André Oriani commented on HDFS-1111:
------------------------------------

[ I started writing  before Rodrigo's second comment]

After resolving HDFS-1031 and having acquired more experience on the problem, I 
think:

1) Saying "Here are a few files that may be corrupted:" is definitely wrong. 
The files are corrupted for sure. The doubt is whether the list is complete or 
not.
2) The output is not sorted. A sorted output would make admin's life easier  
(new Jira?)
3) fsck handles options in a very simple way.  'fsck / -move -delete" is 
accepted although move and delete options are mutually exclusive (new Jira?)
4) Changes here shall be reflected on HDFS-1031 ( add a link to it?)
5) @ Rodrigo : why not returning a struct comprised of the list and a flag to 
say if list is complete. I dunno if it is the best thing to do, but at least it 
will let things more intuitive



> getCorruptFiles() should give some hint that the list is not complete
> ---------------------------------------------------------------------
>
>                 Key: HDFS-1111
>                 URL: https://issues.apache.org/jira/browse/HDFS-1111
>             Project: Hadoop HDFS
>          Issue Type: New Feature
>            Reporter: Rodrigo Schmidt
>            Assignee: Rodrigo Schmidt
>
> If the list of corruptfiles returned by the namenode doesn't say anything if 
> the number of corrupted files is larger than the call output limit (which 
> means the list is not complete). There should be a way to hint incompleteness 
> to clients.
> A simple hack would be to add an extra entry to the array returned with the 
> value null. Clients could interpret this as a sign that there are other 
> corrupt files in the system.
> We should also do some rephrasing of the fsck output to make it more 
> confident when the list is not complete and less confident when the list is 
> known to be incomplete.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to