Clustered Controller States

Clustered controller nodes are in one of the following states at any given time.

Table 2-2 Cluster States

BUI/CLI State Description

BUI/CLI State	Description
Clustering is not configured `AKCS_UNCONFIGURED`	Clustering has not yet been configured. The cluster is either being set up or the cluster setup task has never been completed. The peer controller typically is in state `UNKNOWN` (disconnected or restarting).
Unknown (disconnected or restarting) `UNKNOWN`	The controller is powered off or rebooting, all of its cluster interconnect links are down, or clustering has not yet been configured.
Active (takeover completed) `AKCS_OWNER`	Clustering is configured, and this controller has taken control of all shared resources in the cluster. A controller enters this state in the following cases: Immediately after cluster setup is completed from this controller's user interface. When this controller detects that its peer has failed, for example after a takeover. The controller remains in this state until an administrator manually executes a failback operation. See Cluster Takeover and Failback.
Ready (waiting for failback) `AKCS_STRIPPED`	Clustering is configured, and this controller does not control any shared resources. A controller enters this state in the following cases: Immediately after cluster setup is completed from the user interface of this controller's peer controller. Following a reboot, power disconnect, or other failure. The controller remains in this state until an administrator manually executes a failback operation. See Cluster Takeover and Failback.
Active `CLUSTERED`	Clustering is configured, and both nodes own shared resources according to their resource assignments. If each node owns a ZFS pool and is in the `CLUSTERED` state, then the two nodes form what is commonly called an active-active cluster.
Rejoining cluster ... `REBOOTING`	The appliance has recently rebooted, or the appliance management software is restarting after an internal failure. Resource state is being resynchronized.

Clustering is not configured

AKCS_UNCONFIGURED

Clustering has not yet been configured. The cluster is either being set up or the cluster setup task has never been completed. The peer controller typically is in state UNKNOWN (disconnected or restarting).

Unknown (disconnected or restarting)

UNKNOWN

The controller is powered off or rebooting, all of its cluster interconnect links are down, or clustering has not yet been configured.

Active (takeover completed)

AKCS_OWNER

Clustering is configured, and this controller has taken control of all shared resources in the cluster. A controller enters this state in the following cases:

Immediately after cluster setup is completed from this controller's user interface.
When this controller detects that its peer has failed, for example after a takeover.

The controller remains in this state until an administrator manually executes a failback operation. See Cluster Takeover and Failback.

Ready (waiting for failback)

AKCS_STRIPPED

Clustering is configured, and this controller does not control any shared resources. A controller enters this state in the following cases:

Immediately after cluster setup is completed from the user interface of this controller's peer controller.
Following a reboot, power disconnect, or other failure.

The controller remains in this state until an administrator manually executes a failback operation. See Cluster Takeover and Failback.

Active

CLUSTERED

Clustering is configured, and both nodes own shared resources according to their resource assignments. If each node owns a ZFS pool and is in the CLUSTERED state, then the two nodes form what is commonly called an active-active cluster.

Rejoining cluster ...

REBOOTING

The appliance has recently rebooted, or the appliance management software is restarting after an internal failure. Resource state is being resynchronized.