[
https://issues.apache.org/jira/browse/HDDS-12235?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17924754#comment-17924754
]
Wei-Chiu Chuang commented on HDDS-12235:
----------------------------------------
Currently, there's a 0.01% non-DFS reserved space being considered.
Plus, when importing a container it requires an additional 2* container size of
space (because of storing the temporary tarball too)
However, SCM allows up to 20 concurrent container re-replication happening to
any DataNodes.
So all in all, it should make sure at least 20 * 2 * max container size (5GB) /
number of volumes = 200GB / num(volumes) of volume space is still available.
> Container import processing should respect reserved space at DN
> ---------------------------------------------------------------
>
> Key: HDDS-12235
> URL: https://issues.apache.org/jira/browse/HDDS-12235
> Project: Apache Ozone
> Issue Type: Bug
> Reporter: Uma Maheswara Rao G
> Priority: Major
>
> Concurrent container download and import processing should agree on available
> space at DN accurately. It should always respect du reserve space as well.
> Otherwise, DN volume can be out of disk space.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
