[
https://issues.apache.org/jira/browse/HDFS-13134?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16365248#comment-16365248
]
Lokesh Jain commented on HDFS-13134:
------------------------------------
[~anu] Thanks for reviewing the patch! v3 patch fixes checkstyle issues, find
bugs and unit test failures.
I think we can pass list of containers to the format open container command as
we do not want containers with replication factor one to be deleted. Further we
can use it for deleting over replicated containers.
Also currently in ContainerManagerImpl#init:180 we throw an exception if read
from a single container fails in readContainerInfo call. Should we just move
forward to the next container instead of throwing an exception?
> Ozone: Format open containers on datanode restart
> -------------------------------------------------
>
> Key: HDFS-13134
> URL: https://issues.apache.org/jira/browse/HDFS-13134
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Lokesh Jain
> Assignee: Lokesh Jain
> Priority: Major
> Attachments: HDFS-13134-HDFS-7240.001.patch,
> HDFS-13134-HDFS-7240.002.patch, HDFS-13134-HDFS-7240.003.patch
>
>
> Once a datanode is restarted its open containers should be formatted. Only
> the open containers whose pipeline has a replication factor of three will
> need to be formatted. The format command is sent by SCM to the datanode after
> the corresponding containers have been successfully replicated.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
