Owner's manual

Switchovers
ServerNet Nomadic Disk User’s Guide—425485-001
5-7
Unplanned Switchovers
Unplanned Switchovers
When a primary node fails, you perform an unplanned switchover to continue operations
on the backup node. You must planned for the possibility of an unplanned switchover
before a primary node failure occurs. For an example of how the switch modules support
the disk drives after switchover, see Remote Mirroring After a Switchover
on page 1-4.
Example of an Unplanned Switchover From Primary Node to Backup Node
The primary node \PRI loses all power. The operator at the backup node is notified of
the outage. The following procedure performs an unplanned switchover of the primary
node applications.
From the Primary Node \PRI
Manually toggle the fiber-optic switches at the primary node to the opposite position as
described in Switch Card Toggle Switches on page 4-8.
From the Backup Node \BCK
1. If you are not using Nomadic Disk Manager:
a. Change the fiber-optic switches at \BCK to the opposite position as described in
Moving the Fiber Optic Switches to the Opposite Position
on page 5-4.
b. Use the SWCI STATUS command to verify that switch modules are in the
correct position.
SWCI $SWITCH1.#PRI> STATUS
c. Use the SCF START DISK command to start the nomadic volumes that were
switched from \PRI.
-> START DISK $nomadicvol, $nomadicvol
2. If you are using the Nomadic Disk Manager:
a. Log on as the super ID and starts the Nomadic Disk Manager at a TACL
prompt.
> NOMADIC
b. Use the TAKE command to take ownership of the Nomadic volumes from \PRI:
1 ND> TAKE DATA
The TAKE command toggles the appropriate switch cards, switching the
volumes to \BCK.
3. Start applications that access data on the Nomadic volumes, and continue
processing.