SVM (vserver) stopped during head upgrade procedure in a MetroCluster
Applies to
- MetroCluster
- ONTAP 9.5+
- FC-VI link cross cabled
- mgwd.log containing the follow sequence
has_remote_cluster_done_switchover was retried. Retry count : 9
has_remote_cluster_done_switchover failed with err = FCVI Interconnect might be down. For fabric configurations, use "storage switch add" to start monitoring the switches.
Issue
- During a head upgrade all SVMs are entering a stopped state.
- Vserver start command fails with error:
Failed to determine if the remote cluster has done a switchover. Reason :
Internal error occurred on node "ClusterA-02" in cluster "ClusterA".
Details: (FCVI Interconnect might be down. For fabric configurations, use
"storage switch add" to start monitoring the switches. Then run "storage switch
show -port" to verify that the status of the used port is "online", and run
"storage switch show -san-config" to verify that the fabric configuration
(vsan/zone) is appropriate. For Stretch configurations, verify the FCVI
cabling.).