[HDDS-8448] Send FCR to SCM on volume failure

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.4.0
• Labels: pull-request-available
• Assignee: Ashish Kumar
• Reporter: Arun Sarin
• Created: 18/Apr/23 17:49
• Updated: 18/Nov/24 19:57
• Resolved: 06/Jun/23 13:18

Description

When a volume fails on the datanode, SCM should be notified which containers need to be replicated. Currently the volume failure will mark all containers on the volume unhealthy, and the FCR will be sent at the regular interval. Instead the containers should be removed from the container set and the FCR should be sent immediately so SCM learns of the containers that are no longer present.

Issue Links

• is related to: HDDS-8447 Datanodes should not process container deletes for failed volumes
• is related to: HDDS-11650 ContainerId list to track all containers created in a datanode
• split from: HDDS-8441 [EC] Unhealthy replica not getting replaced by healthy replica when hdds.datanode.dir path is deleted
• links to: GitHub Pull Request #4773