[ https://issues.apache.org/jira/browse/HDFS-2282?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Todd Lipcon updated HDFS-2282:
------------------------------
Attachment: hdfs-2282-20.txt
Here's a patch for 20. Before resolving we should write a unit test and also double check
that this bug doesn't show up in trunk.
> Semi-harmless race between block reports and block invalidation
> ---------------------------------------------------------------
>
> Key: HDFS-2282
> URL: https://issues.apache.org/jira/browse/HDFS-2282
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: data-node
> Affects Versions: 0.20.203.0
> Reporter: Todd Lipcon
> Attachments: hdfs-2282-20.txt
>
>
> In the 0.20 codebase, block reports are not synchronized in any way against mutations
to the actual file structure on disk. If a file is removed from a directory while the block
report is scanning that directory, it will be mistakenly reported as existing with a length
of 0, since File.length() on a non-existent file returns 0.
> This results in an error being logged on the DataNode when the NN sends it a second block
deletion request for the already-deleted block. I believe it to be harmless, but the error
message can concern users.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
|