[
https://issues.apache.org/jira/browse/HDFS-1547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12978047#action_12978047
]
Sanjay Radia commented on HDFS-1547:
------------------------------------
>4. I really like the fact that you are proposing the decommissioned nodes are
>not auto shutdown.
Suresh thinks I paid you say this :-) (I will send you the $10 shortly).
+++++1
One cannot infer that a DN completed its decommission by the fact that it
shutdown. A DN may die for an unknown reason. One has to check the
NN state to see if a DN was successfully decommissioned.
I also agree that since we don't need to support the original motivation for
exclude file; there is one file and is used for decommisioning.
I could live with using the name "excludes" but would prefer to give it the
name "decomissioned"; yes it breaks compatibility but going forward
it will have a clean simple meaning and no one will need read some
documentation which says "excludes" use to mean XXX + decommision and now it
means decommision. Lets name this file "decommision" and say that "excludes" is
not supported.
> Improve decommission mechanism
> ------------------------------
>
> Key: HDFS-1547
> URL: https://issues.apache.org/jira/browse/HDFS-1547
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: name-node
> Affects Versions: 0.23.0
> Reporter: Suresh Srinivas
> Assignee: Suresh Srinivas
> Fix For: 0.23.0
>
>
> Current decommission mechanism driven using exclude file has several issues.
> This bug proposes some changes in the mechanism for better manageability. See
> the proposal in the next comment for more details.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
