[
https://issues.apache.org/jira/browse/HDFS-6911?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14109775#comment-14109775
]
Jing Zhao commented on HDFS-6911:
---------------------------------
The patch looks good to me. +1
Is there anyway to have a unit test for this? It may be difficult to test this
in an end to end test thus it will be helpful if we can write a separate test
for the internal state.
Besides, to keep all the blocks to need to be moved in memory may be a big
burden. Different with Balancer, the blocks to move are identified by scanning
files in the namespace, and only files in snapshots may share the same set of
blocks with different paths. In that case, maybe a more efficient way here is
to track the inode id (retrieved from HdfsFileStatus) for files that contain
blocks for moving?
> Archival Storage: check if a block is already scheduled in Mover
> ----------------------------------------------------------------
>
> Key: HDFS-6911
> URL: https://issues.apache.org/jira/browse/HDFS-6911
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: balancer, namenode
> Reporter: Tsz Wo Nicholas Sze
> Assignee: Tsz Wo Nicholas Sze
> Attachments: h6911_20140823.patch
>
>
> Similar to balancer, Mover should remember all blocks already scheduled to
> move (movedBlocks). Then, check it before schedule a new block move.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
