Question & Answer
Question
How to replace SSP disk with a new disk type.
Cause
I need to remove disks in the SSP and replace them with a new disk type.
Answer
To remove and replace a physical volume in a shared storage pool, you can use
chsp
command.
For details and restrictions see
Replacing physical volumes in the storage pool
before using the following procedure.
Best practice recommendations prior to replacing the SSP disk:
- Ensure the cluster, SSP, and all nodes are in good health.
- Ensure up-to-date backups are in place.
- Ensure the new disk/pv (hdisk5 in the following example) has been configured, and it is available to all VIOS nodes in the cluster. Remember, the new disk must be the same size or larger than the one being replaced.
PROCEDURE
The following procedure is an example from a disk replacement on a 2-node cluster with node1 at ioslevel 2.2.6.23 and node2 at 2.2.6.10 where hdisk6 (oldpv) was replaced by hdisk5 (newpv) and the PV (hdisk#) names match on both nodes. In other words, hdisk6 on node1 is also hdisk6 on node2. This is not a requirement but facilitates things from the VIOS admin perspective.
1. To check cluster/SSP/nodes status, i.e.
$ cluster -status -clustername hammond -verbose
Cluster Name: hammond
Cluster Id: cfe086b2d18c11e880035cf3fcebfe55
Cluster State: OK
Repository Mode: EVENT
Number of Nodes: 2
Nodes OK: 2
Nodes DOWN: 0
Pool Name: hammondsp
Pool Id: 000000000913AF5F000000005BC65C0D
Pool Mirror State: NOT_MIRRORED
Node Name: p7virtviosnode1.dfw.ibm.com
Node Id: cfec1d1ad18c11e880035cf3fcebfe55
Node MTM: 9117-MMB0210F6B1R
Node Partition Num: 4
Node State: OK
Node Repos State: OK
Node Upgrade Status: 2.2.6.23 ON_LEVEL
Node Roles: DBN
Pool Name: hammondsp
Pool Id: 000000000913AF5F000000005BC65C0D
Pool State: OK
Node Name: p7virtviosnode2.dfw.ibm.com
Node Id: b2e4ed7415ec11e9805cf7f6b1b158e2
Node MTM: 9117-MMB0210F6B1R
Node Partition Num: 5
Node State: OK
Node Repos State: OK
Node Upgrade Status: 2.2.6.10 ON_LEVEL
Node Roles:
Pool Name: hammondsp
Pool Id: 000000000913AF5F000000005BC65C0D
Pool State: OK
$
2. Verify the "oldpv" (hdisk6 in this example) is part of SSP:
p7virtviosnode1
$ lspv -clustername hammond -sp hammondsp PV NAME SIZE(MB) PVUDID hdisk6 51200 332136005076802820B10680000000000037604214503IBMfcp hdisk7 51200 332136005076802820B10680000000000037704214503IBMfcp hdisk8 51200 332136005076802820B10680000000000037804214503IBMfcp $
p7virtviosnode2
$ lspv -clustername hammond -sp hammondsp PV NAME SIZE(MB) PVUDID hdisk6 51200 332136005076802820B10680000000000037604214503IBMfcp hdisk7 51200 332136005076802820B10680000000000037704214503IBMfcp hdisk8 51200 332136005076802820B10680000000000037804214503IBMfcp $
You can also use 'pv -list' if you wish output to include the PV STATE.
3. Verify the new disk/pv (hdisk5 in this case) is free/available for use. In this example, hdisk5,9,10,11,12 are reported as free/available on both nodes:
$ lspv -free NAME PVID SIZE(megabytes) hdisk5 none 51200 hdisk9 00f7f6b1d9c4176f 51200 hdisk10 none 51200 hdisk11 none 51200 hdisk12 none 51200 $ lspv -avail NAME PVID SIZE(megabytes) hdisk5 none 51200 hdisk9 00f7f6b1d9c4176f 51200 hdisk10 none 51200 hdisk11 none 51200 hdisk12 none 51200 $
4. Run chsp command to replace the original disk (oldpv) with the new disk (newpv), as shown below. The command can be run from any node in the cluster. In this example, we are replacing oldpv/hdisk6 with newpv/hdisk5, and chsp was ran from the DBN (p7virtviosnode1):
$ chsp -replace -clustername hammond -sp hammondsp -oldpv hdisk6 -newpv hdisk5 Current request action progress: % 5 Current request action progress: % 11 Current request action progress: % 99 Current request action progress: % 100 Current request action progress: % 100 $
5. Verify the disk was replaced, i.e.
$ lspv -clustername hammond -sp hammondsp PV NAME SIZE(MB) PVUDID hdisk7 51200 332136005076802820B10680000000000037704214503IBMfcp hdisk8 51200 332136005076802820B10680000000000037804214503IBMfcp hdisk5 51200 332136005076802820B10680000000000037504214503IBMfcp <- NEW DISK $
Note: lspv output above is the same on p7virtviosnode2, so the SSP disk replacement was automatically propagated to all nodes in the cluster.
At this point, the oldpv/hdisk6 is released and can be removed if it is no longer needed or used for something else. This can be validated as shown below, i.e. output from both VIOS nodes:
$ lspv -free NAME PVID SIZE(megabytes) hdisk6 00f7f6b1d998a5c7 51200 hdisk9 00f7f6b1d9c4176f 51200 hdisk10 none 51200 hdisk11 none 51200 hdisk12 none 51200 $ lspv -avail NAME PVID SIZE(megabytes) hdisk6 00f7f6b1d998a5c7 51200 hdisk9 00f7f6b1d9c4176f 51200 hdisk10 none 51200 hdisk11 none 51200 hdisk12 none 51200 $
The above procedure was done while SSP storage was provisioned to an AIX client for user volume group created on hdisk10 (backed by LU on p7virtviosnode1) mirrored to hdisk11 (backed by LU on p7virtviosnode1) with a file system where file system I/O was validated to remain successful before and after replacing the SSP disk. AIX client errlog was validated to have no new entries after the SSP disk was replaced.
Related Information
[{"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Product":{"code":"SSPHKW","label":"PowerVM Virtual I\/O Server"},"Component":"","Platform":[{"code":"PF002","label":"AIX"}],"Version":"VIOS 2.2","Edition":"","Line of Business":{"code":"LOB57","label":"Power"}}]
Was this topic helpful?
Document Information
More support for:
PowerVM Virtual I/O Server
Software version:
VIOS 2.2
Operating system(s):
AIX
Document number:
795514
Modified date:
20 October 2021
UID
ibm10795514
Manage My Notification Subscriptions