Automatic unplanned switchover in MetroCluster FC configurations
In MetroCluster FC configurations, certain scenarios can trigger an automatic unplanned switchover (AUSO) in the event of a site-wide controller failure to provide nondisruptive operations. AUSO can be disabled if desired.
In a four-node and eight-node MetroCluster FC configuration, an AUSO can be triggered if all nodes at a site are failed because of the following reasons:
Node power down
Node power loss
Node panic
If an AUSO occurs, disk ownership for the impaired node's pool0 and pool1 disks is changed to the disaster recovery (DR) partner. This ownership change prevents the aggregates from going into a degraded state after the switchover.
After the automatic switchover, you must manually proceed through the healing and switchback operations to return the controller to normal operation.
Changing the AUSO setting
AUSO is set to auso-on-cluster-disaster by default. Its status can be viewed in the metrocluster show command.
You can disable AUSO with the metrocluster modify -auto-switchover-failure-domain auto-disabled command. This command prevents triggering AUSO in DR site-wide controller failure. It should be run on both the sites if you want to disable AUSO on both the sites.
AUSO can be reenabled with the metrocluster modify -auto-switchover-failure-domain auso-on-cluster-disaster command.
AUSO can also be set to auso-on-dr-group-disaster . This advance level command triggers AUSO on HA failover at one site. It should be run on both the sites with the metrocluster modify -auto-switchover-failure-domain auso-on-dr-group-disaster command.
The AUSO setting during switchover
When switchover occurs, the AUSO setting is disabled internally because if a site is in switchover, it cannot automatically switch over.
Recovering from AUSO
To recover from an AUSO, you perform the same steps as for a planned switchover.