I have a two node 2008 R2 cluster running two instances of SQL. Each has 4 disks plus there is a quorum disk. All disks are iSCSI connected, the first instance connects to one SAN and the second to another. All works fine, but we need to move the iSCSI luns to a new SAN (single one instead of two separate ones that we have now). The plan is to offline the cluster, disconnect the iSCSI targets, move the LUNS (so all 9 are now on a new single SAN), add the new iSCSI connections to the nodes. The question is, will the MSCS identify these disks as the same ones and therefore we can just bring the instances online and all is done, or will they detect them as new disks and need to configure them, in which case its a much bigger job? Has anyone done this before? The disks will have the serial number copied across but I don't know how the disks are identified inside a cluster - if its just the serial then that should be OK but if it uses more iSCSI information or the LUN numbers then it might be complicated (as the LUN numbers will change).

I have a two node 2008 R2 cluster running two instances of SQL. Each has 4 disks plus there is a quorum disk. All disks are iSCSI connected, the first instance connects to one SAN and the second to another. All works fine, but we need to move the iSCSI luns to a new SAN (single one instead of two separate ones that we have now). The plan is to offline the cluster, disconnect the iSCSI targets, move the LUNS (so all 9 are now on a new single SAN), add the new iSCSI connections to the nodes. The question is, will the MSCS identify these disks as the same ones and therefore we can just bring the instances online and all is done, or will they detect them as new disks and need to configure them, in which case its a much bigger job? Has anyone done this before? The disks will have the serial number copied across but I don't know how the disks are identified inside a cluster - if its just the serial then that should be OK but if it uses more iSCSI information or the LUN numbers then it might be complicated (as the LUN numbers will change). They have different SCSI Product/Vendor IDs so you'll have to re-create the config rather then just copy the data.StarWind iSCSI SAN & NAS

Even if they're the same product and vendor? It's NetApp NAS to NetApp NAS...

The cluster software wirtes a disk signature to each drive (LUN) and that is how it keeps track of them. If you preserve the complete contents of the LUNs, it should not have any issues recognizing them on a different IQN. As with any work on a production cluster, you should test the process in a lab before running it on production. Otherwise your production system has become your lab system..:|:.:|:. tim

Hi, So you are going to replace the old disk resources with a new SAN device. In Windows Server 2003, these is Cluster Server Recovery Utility (ClusterRecovery.exe) which you can use to replace disk resource. But from Windows Server 2008, the ClusterRecovery.exe is no longer available. From Windows Server 2008, disk resource has built-in recovery function, which you can use to replace disk resource. When we use repair function, old disk resource got removed from under the control of the cluster. The mainly steps are: Connect new LUN to cluster; online new disk; initialize disk; and create new volume; format the diskGo into failover cluster manager, change the old disk resource to not restart to failover when simulate a failure of this disk. Right-click the disk resource, [Properties], [Policies Tab]. Set the Response to resource failure to If resources fails, do not restart.Simulate failure of this old disk, then we can use repair function to replace failed disk (that is not really failed). Right-click the disk resource, More actions, Repair. This will launch the Repair a Disk Resource window, then we can select the new added disk resourceBring new resource online, you can change its drive letter to match the old driver letterReset failover policy we changed in step 2 to original statusNow, take the old disk resource online, and copy data to new disk resource.Remove old disk resource. Try that and give us feedback for further troubleshooting, for more information please refer to following MS articles: Replacing a Shared Disk on a 2008 Failover Cluster http://blogs.technet.com/b/askcore/archive/2011/08/26/replacing-a-shared-disk-on-a-2008-failover-cluster.aspx How to move Windows 2008 / SQL 2005 Cluster Resources to New SAN http://social.technet.microsoft.com/Forums/en-US/winserverClustering/thread/d896d789-8dcd-4670-82ca-85d82727728f/ Hope this helps! TechNet Subscriber Support If you are TechNet Subscription user and have any feedback on our support quality, please send your feedback here.Lawrence TechNet Community Support