Change Volumegroup Device Files in a running MC/SG-Cluster

[Rainer Kranner]

Dear All

Do I right, if I do switch Devicefiles from Quickloop to Fabric in a SAN with a running
MC/SG Cluster, then i can switch all Pakages to one Node and make :

1.) switch all pakages to 1th Node
2.) vgexport -p -s -m /tmp/vgxx.map /dev/vgxx
3.) switch SAN Device files on 2nd Node
4.) vgimport -s -m /tmp/vgxx.map /dev/vgxx to
the 2nd Node with new Devicefiles.
5.) switch Pakages back to the 2nd Node
6.) switch SAN Device files on 1th Node
7.) Change Clusterconfig with new
Clusterlockdisk Devicefile

let me know if i wrong or right or correct me.

thanks in advance
Rainer

[Thayanidhi]

Hi,

Your idea sound s good but I don't think it's possible to do so.
Reasons.

After you move the package and modify to Fabric in SAN,
the HW paths will change in all the systems(including the node which
runs all the package) and all the disks will be inaccesible, becuase VG
configuration still pointing to old device files (old HW paths).


You must plan cluster outage to do this job.

TT

[Kent Ostby]

There are several good FAQ documents related to this issue.

This one is called "Consequences of changing Disk Paths in a SG cluster" :

http://www2.itrc.hp.com/service/cki/search.do?category=c0&mode=id&searchString=UMCSGKBRC00006938&searchCrit=allwords&docType=Security&docType=Patch&docType=EngineerNotes&docType=BugReports&docType=Hardware&docType=ReferenceMaterials&docType=ThirdParty&search.x=31&search.y=9

There is also:
Replacing a disk in a ServiceGuard configuration which may lend some data for you.

http://www2.itrc.hp.com/service/cki/search.do?category=c0&mode=id&searchString=UMCSGKBRC00007765+&searchCrit=allwords&docType=EngineerNotes&search.x=26&search.y=4

Best regards,

Kent M. Ostby

[Colin Topliss]

Almost.

But you will need to re-apply the cluster configuration because of the change of device file for the clusterlock volume. That, unfortunately, requires an outage.

You will find that the cluster will fail over quite correctly even if you *don't* apply the new configuration. The trouble is that it will be unstable, and when you come to restart your cluster at a later point in time it will fail.

Just been through that little game myself!

Col.

[Rajeev Shukla]

This is the way i migrated from quickloop to fabric in MC/SG environment.

1. Prepare a list of ioscan on all nodes and copy of /etc/lvmtab (keep a backup)
2. export all volume groups in shared mode on all nodes in the cluster (vgexport -s -v -p -m ...)
3. Shutdown the cluster.
4. change all nodes from quickloop to fabric.
5. vgimport all volume groups in shared mode (you will have to export all volume groups from /etc/lvmtab file if in previous step you used -p)
6. create a map of all device files with refrence to old ioscan and new ioscan on all nodes.
7. get the device id of new cluster lock disk on all nodes
8. make changes in cluster config file and cmcheckconf and then cmapplyconf

As far as i remember, this should move you'r whole cluster and all nodes to fabric