[
https://issues.apache.org/jira/browse/HDDS-11463?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17882142#comment-17882142
]
Ethan Rose commented on HDDS-11463:
-----------------------------------
This would be a good thing to add. Datanodes already report disk health to SCM
via [storage
reports|https://github.com/apache/ozone/blob/6c6dc4352e475c56a5972f8fe355c0aca2a6554c/hadoop-hdds/container-service/src/main/java/org/apache/hadoop/ozone/container/common/impl/StorageLocationReport.java#L39]
attached to heartbeats. We might need to expose this to users better. I think
Recon already has a page for this, but an SCM CLI in json format would be good
as well.
> Track and display failed DataNode storage locations in SCM.
> -----------------------------------------------------------
>
> Key: HDDS-11463
> URL: https://issues.apache.org/jira/browse/HDDS-11463
> Project: Apache Ozone
> Issue Type: Sub-task
> Components: Ozone Datanode
> Reporter: Shilun Fan
> Assignee: Shilun Fan
> Priority: Major
>
> We currently frequently encounter under-replicated containers, which is
> usually due to issues such as DataNode heartbeat failures or damaged disks on
> DataNodes leading to container loss and subsequent automatic recovery.
> However, it is challenging to identify damaged disks in a large cluster. This
> JIRA will allow DataNodes to actively report damaged disks to the SCM, which
> can then display the damaged disks in a list format.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
