Removing and replacing the frame assembly

This topic describes how to replace the SAN Volume Controller 2145-8F2 and SAN Volume Controller 2145-8F4 frame assembly.

The SAN Volume Controller 2145-8F2 and SAN Volume Controller 2145-8F4 frame assembly must be replaced when the system board fails or when replacing other system board components does not isolate the error.

Note: If you are replacing the frame assembly after you replaced the service controller, you must ensure that the node can be added back to the cluster. Use the following choices, which are based on the actions that you must take or have already taken, to determine how the node can be added back to the cluster:

If you are replacing the node because the new service controller did not fix the original problem and you could not update the WWNN of the service controller, use the original service controller when you perform the frame replacement. This ensures that the WWNN is correct.
If you have already made any required updates to the worldwide node name (WWNN) and the updates are stored in the service controller, continue with the following instructions to replace the frame assembly.
If you must use a new service controller when you replace the frame assembly, use the following instructions to replace the frame assembly and make sure that you follow the special instructions in step 11.

Perform the following steps to replace the frame assembly:

Write down the 7-character serial number from the serial number label that is on the front of the node.

For the SAN Volume Controller 2145-8F2 and SAN Volume Controller 2145-8F4 models only: If you cannot read the serial number or if you want to verify that it is correct, you can also find it on the node vital product data (VPD).

Perform the following steps to find the serial number in the VPD:
1. Start the SAN Volume Controller Console application.
2. Display the VPD for the failed node.
3. Note the system serial number.
Turn off the node while ensuring that its data is mirrored and synchronized. See MAP 5350: Powering off a SAN Volume Controller node for more information.
Remove the node from the rack and place it on a flat, static-protective surface. See Removing the SAN Volume Controller from a rack.
Remove the top cover. See Removing the top cover.
Remove the fibre-channel adapter cards from the frame assembly that you are replacing.
Remove the service controller from the frame assembly you are replacing.
Install the fibre-channel adapter cards into the new frame assembly.
Install the service controller into the new frame assembly.
Replace the top cover. See Replacing the top cover.
Place the node in the rack. See Replacing the SAN Volume Controller in a rack.
Start the node using the node rescue procedure.
Note: If you have to install a new service controller in the frame assembly, you must perform the next steps to allow the node rescue to find a suitable donor node.
1. Turn on the node. Use your SAN monitoring tools to determine the WWNN of the node.
2. Rezone the switch to allow at least one port of the replaced node to connect to at least one port of an existing node.
3. Perform a node rescue.
4. Set the WWNN of the replacement node to match the node that it replaced. See Node WWNN for more information.
5. Rezone the switch to its original settings.
6. Add the node back into the cluster.
7. Clearly label the original service controller to indicate that its WWNN has been re-used, and that the part cannot be used without first changing the WWNN.
Attention: It is essential that you perform all the stages of the next step to ensure that the replacement system is set to the serial number of the original machine.
If you are performing this repair as part of a directed maintenance procedure, you will be prompted to type the machine serial number that you noted above. Otherwise, perform the following steps:
1. Delete the failed node from the cluster.
2. Add the repaired node to the cluster.
3. Start the command-line interface (CLI). See Accessing the SAN Volume Controller CLI.
4. Issue the following command:
  svctask writesernum -sernum nodeserialnumber nodename
  
  where nodeserialnumber is the number that you noted previously and nodename is the name of the repaired node that you added in this step.
  For example, to write the machine serial number to the system board when the serial number is "13-FEDCB" and the node name is "ZYXW3," you would issue the following command:
```
svctask writesernum –sernum 13FEDCB ZYXW3
```
  Note: The node will restart as soon as the serial number has been written to it.
5. Write the serial number that you noted previously on the blank serial number label on the front of the node.