Hello baszek,
Assuming as you mentioned "HDD" that this is a Hybrid configuration and cannot not have Dedupe enabled (meaning just this disk is failed, not the whole disk-group).
Check the current Object configuration status:
# cmmds-tool find -f python | grep CONFIG_STATUS -B 4 -A 6 | grep 'uuid\|content' | grep -o 'state\\\":\ [0-9]*' | sort | uniq -c
This *should* return all Object as config-state 7 (e.g. Healthy)
"My question is - do I have to put Vsan host to Maintenance mode before deleting failed drive from disk group ? or whole replacement can be done online ?"
Provided your controller and the mode it is using supports hot-add/remove then replacing the disk live shouldn't cause any issues.
While it is not mandatory to put a host in Maintenance Mode before removing failed disks, however if you are unfamiliar with confirming the availability state of the data, this can be performed temporarily (e.g. with 'Ensure Accessibility').
Otherwise removing the disk with the correct options without MM will suffice - removing the failed disk with 'Ensure Accessibility' option should be feasible if the data on the cluster is otherwise healthy.
Bob
