Removing a command device from the UCP Advisor VM or the remote gateway VM

Unified Compute Platform (UCP) Advisor Administration Guide

Version

4.6.x

Audience

anonymous

Part Number

MK-92UCP119-15

ft:lastEdition

2024-09-24

There might be occasions when you want to remove a command device from the UCP Advisor VM or the remote gateway VM, such as when resolving storage configuration issues or when decommissioning a storage system. You can remove the command device even when it is disconnected from the UCP Advisor VM or the remote gateway VM.

Using SSH, log on to the UCP Advisor VM or the remote gateway VM as ucpadmin.

Verify the existing command devices. Enter:

lsscsi -i

Sample output:

[ucpadmin@UCPAdvisorMasterHost105 ~]$ lsscsi -i
[0:0:0:0] disk VMware Virtual disk 2.0 /dev/sda -
[0:0:1:0] disk VMware Virtual disk 2.0 /dev/sdb -
[0:0:2:0] disk HITACHI OPEN-V-CM 5001 /dev/sdc 360060e80282742005080274200000049
[0:0:3:0] disk HITACHI OPEN-V-CM 5001 /dev/sdd 360060e8028273e005080273e0000000a
[0:0:4:0] disk HITACHI OPEN-V-CM 5001 /dev/sde 360060e8028273d005080273d00000008
[0:0:5:0] disk HITACHI OPEN-V-CM 5001 /dev/sdf 360060e80233abc0050703abc00000024
[0:0:6:0] disk HITACHI OPEN-V-CM 5001 /dev/sdg 360060e80212b1f0050602b1f0000089e
[0:0:8:0] disk HITACHI OPEN-V-CM 9301 /dev/sdh 360060e8012b3d7005040b3d70000019c

This command also displays the disk device and the storage details.

Identify the storage system that corresponds to the command device that you want to remove.
1. In the command output, note the hexadecimal number. These are the four digits after the eighth digit from the right as highlighted in the above sample output.
  For example, 2742 is the hexadecimal number for the /dev/sdc command device.
2. Convert the hexadecimal number to its decimal equivalent.
  For example, the decimal equivalent of 2742 is 10050. This number represents the last five digits of the storage serial number.
3. Use this decimal number to identify the storage device.
Note the command device or devices that you want to remove.
Note: If you are removing the last disk or disks, continue to step 5. However, if you are removing the first disk or the intermediate disks, you must:
1. Remove the RDM disk or disks from the UCP Advisor VM or the remote gateway VM.
2. Reboot the VM.
3. Repeat the procedure from step 1.

Back up the existing sdi-gateway-block deployment. Enter:

kubectl get deploy sdi-gateway-block -n ucp -oyaml > orig-sdi-block-deploy.yaml

Edit the sdi-gateway-block deployment. Enter:

kubectl edit deploy sdi-gateway-block -n ucp

Navigate to the spec > template > spec > containers > 0 > volumeDevices section.

Delete the devicePath and the name lines for each command device.

Before deleting:

volumeDevices:
- devicePath: /dev/sdc
name: dev-sdc
- devicePath: /dev/sdd
name: dev-sdd
- devicePath: /dev/sde
name: dev-sde

After deleting:

volumeDevices:
- devicePath: /dev/sdc
name: dev-sdc
- devicePath: /dev/sdd
name: dev-sdd

Note: The devicePath and the name lines are deleted for the sde command device.

Navigate to the spec > template > spec > volumes section.
Delete the name, persistentVolumeClaim, and claimName lines for each command device.
Before deleting:
```
volumes:
- name: log-dir
persistentVolumeClaim:
claimName: pvc-logs
- name: dev-sdc
persistentVolumeClaim:
claimName: pvc-sdc
- name: dev-sdd
persistentVolumeClaim:
claimName: pvc-sdd
- name: dev-sde
persistentVolumeClaim:
claimName: pvc-sde
```
After deleting:
```
- name: log-dir
persistentVolumeClaim:
claimName: pvc-logs
- name: dev-sdc
persistentVolumeClaim:
claimName: pvc-sdc
- name: dev-sdd
persistentVolumeClaim:
claimName: pvc-sdd
```
Note:
- The name, persistentVolumeClaim, and claimName lines are deleted for the sde command device.
- If you have removed any command devices, you must also remove the devicePath and name lines from the volumeDevices section, and name, persistentVolumeClaim, and claimName lines from the volumes section for each command device. Follow steps 6 to 12.
Exit the edit mode. Press the Esc key.
Save the changes. Enter:
```
:wq
```
Verify the pod status.Enter:
```
kubectl get pods -n ucp -w
```
The old pod must be deleted and the new pod must be created. If the old pod is not deleted or if it is in the Terminating state, delete the old pod forcefully. Enter:
```
kubectl delete pods <POD-NAME> -n ucp --grace-period=0 –force
```
Remove the RDM disk from the UCP Advisor VM or the remote gateway VM.
Note: You can ignore this step if you have already removed the RDM disk while removing the intermediate command devices.