vSAN objects stuck in reduced availability with no rebuild after converting a standard cluster to stretched
Article ID: 417068
Updated On:
Products
Issue/Introduction
Depending on the number of hosts in the standard cluster and the storage policies being used, it's possible for storage policies to become incompatible with a stretched cluster configuration due to the change in fault domains.
Example: A 5-node standard cluster will have 5 total fault domains (one for each host), if this is converted to a 10-node stretched cluster after adding 5 additional nodes, then there will now be only 3 fault domains (one for each host site + the virtual witness host). A standard cluster RAID5 storage policy will be incompatible with a cluster that has only 3 fault domains as a healthy RAID 5 object requires 4 fault domains, so in this scenario the storage policy needs to be changed to a site mirrored RAID5 to maintain compatibility, otherwise vSAN cannot resync objects into a healthy state due to the lack of 4 fault domains.
Environment
8.0
Cause
Incompatible vSAN storage policy
Resolution
Reconfigure all impacted VMs with a compatible site mirrored or other stretched cluster appropriate storage policy
Check the compatibility of all applicable VM storage policies:
Check the datastore default storage policy for the cluster:
Feedback
