HP Secure Path for Windows

[Shearill Brown]

I have a Microsoft Cluster thats is an active\Passive configuration I am using Veritas Volume manager to manange my volumes and Secure Path for Windows 4.0 SP1. A lot of cooks huh?

On my cluster Server1(Active) and Server2(Passive). Each server has dual Emulex 9002L HBA's. I set the Secure path policy to SQST on the passive node and tried to fail my Microsoft SQL Cluster over from active and passive and the disk group failed. Originally the policy was set to NLB. What is the best policy to use using Secure path? I am trying to just set up redundancy so that when an HBA is pulled from the box the second HBA takes over.

Any Help with Secure path would be great. I am new to this software.

[Steven Clementi]

Shearill:

You did not mention the type of storage array you are using. It would be helpful to know which it is.

There are a few reason why the disk group failed to move over and at this point, it's only speculation as to why without additional information.

1. Are your SAN switches properly configured with zoning? (if configured at all)?

2. Does the cluster group failover properly?

3. Which storage array are you using?

These are only a few of the questions I can ask. Can you address them and/or provide more information?


Happy Holiday's,

Steven

[Richard Perez]

Have you tried to force the failover doing a shutdown of the active node?

Do you see the disks groups as Veritas"imported" or "deported" in the passive node?

What version of Veritas are you using?

[Shearill Brown]

Hello,

My storage Array is an HP1024 and I am using Storage Foundations 4.0 for Veritas volume manager. The Cluster Group fails over fine but the in the SQL group the VM Disk fail. The disk stay on the active node after trying to fail and they work fine on that. I need to failover however without any errors. Is SQST not set up right or not the right Policy to use?

[Richard Perez]

Shearill
try shutting down the active node. This way you will discard "SCSI contention" that might be avoiding a graceful failover of the resources.

If this enables the passive node to own the resources you should try step number 2: create a disk on the XP, present it to both nodes, create a disk group and volume under Veritas and then deport/import the volume with Veritas commands or GUI.