[
https://issues.apache.org/jira/browse/HDDS-140?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16500494#comment-16500494
]
Xiaoyu Yao commented on HDDS-140:
---------------------------------
Thanks [~hanishakoneru] for working on this. The patch looks pretty good to me
overall. I just have few minor comments:
VolumeInfo.java
Line 90-94: this can be changed to below like we do for the
FileIoProvider#mkdirsWithExistsCheck. I wish we could have something similar to
the FileIoProvider for HDDS so that the profiling and I/O checking will be much
easier in future.
Boolean succeeded = rootDir.isDirectory() || rootDir.mkdirs();
Line 124-131:
After shutdown is called, the volume is changed from failed state to
non_existent state here, I have some question wrt. in VolumeSet.java Line
178-182 below.
In VolumeSet.java
Line 178-182: we are putting a volume of non_existent state into a failed
volume map. Should we keep the volume in failed state during the
volumeInfo.failVolume()?
Line 221-223: should we catch exception during the individual volume shutdown
so that if one volume shutdown exception won't leave other df/du threads
lingering around?
> Add DU usage to VolumeInfo
> --------------------------
>
> Key: HDDS-140
> URL: https://issues.apache.org/jira/browse/HDDS-140
> Project: Hadoop Distributed Data Store
> Issue Type: Sub-task
> Reporter: Hanisha Koneru
> Assignee: Hanisha Koneru
> Priority: Major
> Fix For: 0.2.1
>
> Attachments: HDDS-140-HDDS-48.001.patch
>
>
> We need to start a DU thread per VolumeInfo to get the usage info.
> Each VolumeInfo object will wrap a VolumeUsage object to collect usage info.
> The proposed VolumeUsage class will replace ContainerStorageLocation.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
