[
https://issues.apache.org/jira/browse/HDFS-7374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14200625#comment-14200625
]
Zhe Zhang commented on HDFS-7374:
---------------------------------
This use case has a (slightly) conflicting objective as that of HDFS-6791. I
can think of 2 options to accommodate both scenarios:
# Set a timeout (e.g., 10 minutes) limiting the time that a dead DN can stay in
DECOMMISSION_INPROGRESS state.
# If a DN is already dead when decomm starts, indicating that a user is
intentionally decommissioning a dead node, we should allow it to enter
decommission complete state.
[~mingma] and [~jingzhao] please advise if they look reasonable to you, and
whether you prefer one over the other. Or any other approaches. Thanks!
> Allow decommissioning of dead DataNodes
> ---------------------------------------
>
> Key: HDFS-7374
> URL: https://issues.apache.org/jira/browse/HDFS-7374
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: Zhe Zhang
> Assignee: Zhe Zhang
>
> We have seen the use case of decommissioning DataNodes that are already dead
> or unresponsive, and not expected to rejoin the cluster.
> The logic introduced by HDFS-6791 will mark those nodes as
> {{DECOMMISSION_INPROGRESS}}, with a hope that they can come back and finish
> the decommission work. If an upper layer application is monitoring the
> decommissioning progress, it will hang forever.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
