Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I need foolproof instructions for creating a clustered file share on an existing 2-node SQL Server Failover cluster (Windows Server 2003). The scenario is that I have a non-cluster aware application installed on both nodes which needs to create and copy files between subfolders in particular directory. I need a file share that appears to this app as if it's a single folder (e.g. like a network share) regardless of which node in the cluster is active.

I've tried following the instructions in this TechNet article but ran into difficulty straight away. For example, step 1 in the TechNet article says:

  1. Using Windows Explorer, create a folder on the cluster disk

How do I know which disk is the cluster disk?

In step 4 of the same article, you have to create a physical disk resource. I stopped on page 2 of the wizard when it referred to "possible owners" - should this always be both nodes, and what is the effect of ticking the box "Run this resource in a separate Resource Monitor".

And finally, once I've created the file share, how is this exposed to my application? Do I pass it a UNC path, like a regular file share on a network?

Basically, I need an idiots guide so that I don't mess up the existing SQL configuration.

share|improve this question
add comment

1 Answer 1

up vote 1 down vote accepted

These instructions may be easier to follow.

How do I know which disk is the cluster disk?

If you have a 2 node SQL cluster, then you already have a few clustered disk resources. Start cluadmin, if you look inside your resource groups you'll see some physical disks in there under the cluster and SQL group(s). I dont recommend using same disk resources that your SQL Server cluster is using for performance/dependency reasons. The neatest way to do this is to create a new resource group, with a new IP, network name & disk resource. You'll need to provision another disk and/or disk partition that'll be big enough to store the files for your application if you want to do this. How do you do this? it depends on the share storage technology your server is using? It'll be something like a SAN with iSCSI or HBA's or maybe even a directly attached storage array (old school).

Worst case scenario you can probably get away with using your quorum drive that's in the cluster group! it also helps that you have a network name and IP address resources already setup. This will be the hostname in the UNC path you use to access the share.

So, using the instructions I've provided, in step 4. You can use the cluster group! Should be ok from there on?

what is the effect of ticking the box "Run this resource in a separate Resource Monitor?

The resource monitor is a process (resrcmon.exe) that will check if resource is available & if not, lets the cluster service know. If you tick the box, then you'll end up with a seperate resrcmon.exe for just that resource. This may be useful if another resource is causing the monitor to crash, under normal circumstances I'd leave it unticked.

share|improve this answer
    
@Nick Kavadias: Can I reuse the same IP and network name, but just create a new disk resource? –  Jazza Feb 25 '10 at 13:22
    
in the cluster group? sure. just be aware that your disk will follow around the group it's in. –  Nick Kavadias Feb 25 '10 at 13:39
    
Cool - thanks a million! –  Jazza Feb 25 '10 at 14:50
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.