[
https://issues.apache.org/jira/browse/HDFS-6133?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14328914#comment-14328914
]
Yongjun Zhang commented on HDFS-6133:
-------------------------------------
HI [~cmccabe],
My understanding is that HDFS-4420's approach is a passive "pinning": it tells
balancer not to move the blocks belonging to a specified file path; on the
other hand, HDFS-6133 proactively pin the blocks at write time by specifying
favored nodes. I think if we do passive "pinning" after the the blocks of a
file have been moved by balancer, then it may not be preferred behavior. Hi
[~zhaoyunjiong], would you please share your thoughts when you started with
this approach when HDFS-4420's patch was already there?
Hi [~szetszwo] and [~zhaoyunjiong], about HDFS-6133, I was wondering whether
there is any compatibility issue when we do rolling upgrade? There are API
changes at {{DataTansferProtocol#writeBlock}}, what would happen if DNs with
HDFS-6133 (trying to pin) talks to DNs without the fix? Would you please
comment?
Thanks.
> Make Balancer support exclude specified path
> --------------------------------------------
>
> Key: HDFS-6133
> URL: https://issues.apache.org/jira/browse/HDFS-6133
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: balancer & mover, datanode
> Reporter: zhaoyunjiong
> Assignee: zhaoyunjiong
> Fix For: 2.7.0
>
> Attachments: HDFS-6133-1.patch, HDFS-6133-10.patch,
> HDFS-6133-11.patch, HDFS-6133-2.patch, HDFS-6133-3.patch, HDFS-6133-4.patch,
> HDFS-6133-5.patch, HDFS-6133-6.patch, HDFS-6133-7.patch, HDFS-6133-8.patch,
> HDFS-6133-9.patch, HDFS-6133.patch
>
>
> Currently, run Balancer will destroying Regionserver's data locality.
> If getBlocks could exclude blocks belongs to files which have specific path
> prefix, like "/hbase", then we can run Balancer without destroying
> Regionserver's data locality.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
