User`s guide

because, after setting failover, you will not be accessing those units anyway. (The
target controller will only access the copied units.)
See Section 5.2.5 for a detailed description for setting configuration parameters
for dual-redundant configurations.
HSZ Array Controllers: Observe the following considerations when setting
dual-redundant HSZ array controllers for failover:
• Subsystem performance will be better if you balance the assignment of target
IDs across your dual-redundant pair.
• The controller you are copying configuration information to will automatically
restart after entering the SET FAILOVER command.
• You MUST connect both controllers to the same host SCSI–2 bus. If you
connect the controllers in a dual-redundant pair to different host buses, and
one controller fails, the subsequent failover process will cause adverse effects
on your system.
2.3.2 Exiting Failover
If you want to take two controllers out of the failover configuration and both
controllers are still functioning, you must enter the following command:
CLI> SET NOFAILOVER
The OTHER controller will shut down when this command is issued.
2.3.3 Taking Controllers Out of Failover
After one controller in your dual-redundant pair has failed, the surviving
controller services your entire configuration. To take the surviving controller out
of the failover configuration, enter the following command:
CLI> SET NOFAILOVER
You must consider the following before entering the SET NOFAILOVER command
on a surviving controller:
• You do not need to SET NOFAILOVER, unless you must make configuration
changes before you can replace the failed controller.
• If the surviving controller and its cache module are functioning normally, you
can SET NOFAILOVER without special preparation.
• If the surviving controller is running with low write-back cache batteries
(enter the SHOW THIS_CONTROLLER command to check battery status),
you should replace/recharge the batteries before entering SET NOFAILOVER
and before removing the failed controller’s cache module.
Entering SET NOFAILOVER removes the surviving controller from the failover
configuration. You may now make the configuration changes under the surviving
controller. Always SET NOFAILOVER before deleting any units, storagesets, or
single devices.
In a dual-redundant configuration, both controllers have copies of the same
configuration information. Even if you configure disk drives, tape drives,
storagesets, and units after entering the SET FAILOVER command, the
configuration information will be communicated to the second controller when
it is returned to service and the controllers are restored to a dual-redundant
configuration.
2–18 Controller Technical Description