[
https://issues.apache.org/jira/browse/HDFS-4883?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13688562#comment-13688562
]
Brandon Li commented on HDFS-4883:
----------------------------------
Thanks for the patch!
A few comments:
1. in ClientNamenodeProtocolServerSideTranslatorPB.java, req.getFileId() should
be "req.hasFileId() ? req.getFileId() : INodeId.GRANDFATHER_INODE_ID" for
backward compatibility.
2. in ClientNamenodeProtocol.proto, instead of:
required uint64 fileId = 3;
optional ExtendedBlockProto last = 4;
it should be:
optional ExtendedBlockProto last = 3;
optional uint64 fileId = 4;
3. checkLease() throws exception when src and fileId don't match. It seems a
better approach if fileId presents, the namenode API should use fileId instead
of src to serve the request. I am OK if you think this is beyond this JIRA and
doesn't have to be fixed here.
4. It would be nice to add simple unit test to validate the change.
> complete() should verify fileId
> -------------------------------
>
> Key: HDFS-4883
> URL: https://issues.apache.org/jira/browse/HDFS-4883
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: namenode
> Affects Versions: 2.1.0-beta
> Reporter: Konstantin Shvachko
> Assignee: Tao Luo
> Attachments: HDFS-4883.patch
>
>
> It seems inconsistent that complete() avoids verifying real fileId, while
> addBlock() does it meticulously.
--
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
