[
https://issues.apache.org/jira/browse/HDFS-6727?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Lei (Eddy) Xu updated HDFS-6727:
--------------------------------
Attachment: HDFS-6727.003.patch
[~cmccabe] Thank you very much for the detailed review. It helps a lot!
I'v updated the patch to addressing your comments.
bq. It seems like we should add any new block pool objects before we add the
volumes that reference them?
I changed the behavior to loading replica maps first in
{{FsDatasetImpl#addVolumeAndBlockPool}.
bq. This brings up an important point: are we going to allow users to add
directories that already contain block files and so forth in them?
We expect the user can either add a new empty volume or add a volume that moved
from a backup machine. So that the volume should either be empty or filled data
with the same layout version. Thus I've added a new {{StartupOption.HOTSWAP}}
for this case.
I've also changed the error handling to handle the cases of adding volumes to
{{FsDatasetImpl}}. The logic of adding volumes to {{DataStorage}} is on another
JIRA (HDFS-7035), which can be incorporate later.
Would you mind to give another reivew, [~cmccabe].
> Refresh data volumes on DataNode based on configuration changes
> ---------------------------------------------------------------
>
> Key: HDFS-6727
> URL: https://issues.apache.org/jira/browse/HDFS-6727
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: datanode
> Affects Versions: 2.5.0, 2.4.1
> Reporter: Lei (Eddy) Xu
> Assignee: Lei (Eddy) Xu
> Labels: datanode
> Attachments: HDFS-6727.000.delta-HDFS-6775.txt, HDFS-6727.001.patch,
> HDFS-6727.002.patch, HDFS-6727.003.patch, HDFS-6727.combo.patch
>
>
> HDFS-1362 requires DataNode to reload configuration file during the runtime,
> so that DN can change the data volumes dynamically. This JIRA reuses the
> reconfiguration framework introduced by HADOOP-7001 to enable DN to
> reconfigure at runtime.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
