[
https://issues.apache.org/jira/browse/HDDS-7728?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17764748#comment-17764748
]
Stephen O'Donnell commented on HDDS-7728:
-----------------------------------------
For over replicated, I am not sure we need to do anything.
1) The container is over replicated and the delete blocks arrive and are sent
to more than 3 replicas that is ok.
2) If the container gets deleted on the DN first, while the delete blocks is in
flight, then the DN needs to handle that and give an OK response to SCM,
indicating the blocks have been deleted.
Note that with under-replication, its a very similar problem to the balancer.
The balancer causes over replication, and it could create a new replica that
misses out on deletes. And then we would not want to delete that new replica
with the missing deletes, as it was the one the Balancer just created.
Perhaps rather than under-replication, we should be checking for pending
operations against the container (stored in ContainerReplicaPendingOps), but
even then, I suspect there are edge cases that will slip through around over /
under / mis replication and balancing that will leave some oprhan blocks. And
these occassions should be rare enough that we don't want to overly complicate
things - especially if we believe there needs to be an "orphan block" cleanup
task driven by recon anyway.
> Block should be safely deleted from the containers if they are instructed
> from OM and containers are in missing state.
> ----------------------------------------------------------------------------------------------------------------------
>
> Key: HDDS-7728
> URL: https://issues.apache.org/jira/browse/HDDS-7728
> Project: Apache Ozone
> Issue Type: Improvement
> Components: SCM
> Affects Versions: 1.3.0
> Reporter: Uma Maheswara Rao G
> Assignee: Ashish Kumar
> Priority: Major
>
> Currently when OM instructs to delete the blocks and if containers are in
> missing state, deletion may not be processed properly. This Jira to track
> this requirement and implement to safe deletion os blocks what ever state
> they are on. Otherwise containers would never get cleaned up even though all
> blocks in that files deleted.
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
