[ 
https://issues.apache.org/jira/browse/HDFS-4305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13627856#comment-13627856
 ] 

Kihwal Lee commented on HDFS-4305:
----------------------------------

We had a user who accidentally set the block size to something small in his big 
job. That job alone added 10s of millions of blocks when it was still in early 
stage. The blocks/file was large but did not grow large enough to cause trouble 
(0.23 was without HDFS-4304) before taking an admin action, since the progress 
was so slow. The blocks were allocated and completed so frequently that the 
namenode couldn't keep up, let alone the pressure on the namespace and heap. 

In cases like this, max blocks per file may not be useful for detecting the 
issue, but enforcing min block size will help. 

                
> Add a configurable limit on number of blocks per file, and min block size
> -------------------------------------------------------------------------
>
>                 Key: HDFS-4305
>                 URL: https://issues.apache.org/jira/browse/HDFS-4305
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: namenode
>    Affects Versions: 1.0.4, 3.0.0, 2.0.2-alpha
>            Reporter: Todd Lipcon
>            Assignee: Andrew Wang
>            Priority: Minor
>         Attachments: hdfs-4305-1.patch
>
>
> We recently had an issue where a user set the block size very very low and 
> managed to create a single file with hundreds of thousands of blocks. This 
> caused problems with the edit log since the OP_ADD op was so large 
> (HDFS-4304). I imagine it could also cause efficiency issues in the NN. To 
> prevent users from making such mistakes, we should:
> - introduce a configurable minimum block size, below which requests are 
> rejected
> - introduce a configurable maximum number of blocks per file, above which 
> requests to add another block are rejected (with a suitably high default as 
> to not prevent legitimate large files)

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to