User`s guide

Failover should normally complete in less than 10 seconds. If drive I/O is in
progress at the time of failure, the surviving controller must reset any SCSI–2
buses with outstanding I/O.
Whenever you need to revive the controller that was disabled, you must enter the
following command from a terminal connected to the functioning controller:
CLI> RESTART OTHER_CONTROLLER
Then, press the reset (//) button to initialize the controller to be revived.
2.3.4 Using Failover Commands When Write-Back Cache Is in Use
Failover commands can be used while write-back caching is enabled provided the
following restrictions are observed:
• When setting dual-redundant controllers for failover, make sure the target
controller (the controller you are copying configuration data to) has no cache
errors or unwritten cached data. Delete any units on the target controller
to verify there are no cache errors, before entering the SET FAILOVER
command.
• Do not take a dual-redundant controller pair out of failover (SET
NOFAILOVER) with unwritten cached data present in the write-back cache
modules. Doing so will destroy data. Use the SHOW THIS_CONTROLLER
and the SHOW OTHER_CONTROLLER commands to confirm that cache data
has been written. Refer to Section 6.4 for more about unwanted unwritten
cache data in the write-back cache module.
2.3.5 Resolving a Configuration Mismatch after a Hardware Mismatch
The controller configuration is stored in nonvolatile memory on the controller
module. Therefore, when one controller module is substituted for another (as
might occur when replacing a failed controller), the new module’s configuration
contents will not match that of the surviving running controller. When the
replaced controller is restarted, the configuration discrepancy is detected and a
suitable error message is reported.
When a configuration mismatch is detected, the restarted controller will not
recognize or access any devices. It is essentially disabled until the configurations
are made the same.
To resolve a configuration mismatch, enter the following command on the
surviving controller (configuration source) with the known good configuration:
CLI> SET FAILOVER COPY= THIS_CONTROLLER
This will copy the known good configuration to the newly restarted other
controller.
Configuration mismatch also can occur when a subsystem that has never been
configured is started for the first time. The two controller nonvolatile memories
can contain different configurations left over from their last use.
To resolve a mismatch when neither controller contains a known good
configuration, examine the two available configurations, chose one, and issue
the SET FAILOVER COPY=THIS_CONTROLLER command to that controller
to replicate the configuration on the other (the maintenance terminal must be
plugged into the controller with the good configuration information).
Controller Technical Description 2–19