[
https://issues.apache.org/jira/browse/HDFS-9426?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15015437#comment-15015437
]
Vinayakumar B edited comment on HDFS-9426 at 11/20/15 10:33 AM:
----------------------------------------------------------------
bq. I am just wondering whether it will be worth doing more code changes to
avoid introducing the redundant field in the new protocol. Vinayakumar B, what
do you think?
I am thinking current patch is sufficient for the fix. At-least to resolve this
blocker for 2.7.2.
Is there any other way, except this fix and reverting HDFS-7645?
was (Author: vinayrpet):
bq. I am just wondering whether it will be worth doing more code changes to
avoid introducing the redundant field in the new protocol. Vinayakumar B, what
do you think?
I am thinking current patch is sufficient for the fix. At-least to resolve this
blocker for 2.7.2.
Is there any other way, this fix and reverting HDFS-7645?
> Rollingupgrade finalization is not backward compatible
> ------------------------------------------------------
>
> Key: HDFS-9426
> URL: https://issues.apache.org/jira/browse/HDFS-9426
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: Kihwal Lee
> Priority: Blocker
> Attachments: HDFS-9426.branch-2.7.poc.patch, HDFS-9426.trunk.poc.patch
>
>
> After HDFS-7645, the namenode can return non-null {{rollingUpgradeInfo}} in
> heatbeat reponses. 2.7.1 or 2.6.x datanodes won't finalize the upgrade
> because it's not null.
> NN might have to check the DN version and return different
> {{rollingUpgradeInfo}}.
> HDFS-8656 recognized the compatibility issue of the changed semantics, but
> unfortunately did not address the semantics of the heartbeat response.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
