MS Clusters works in a kind of active-passive arrangement .. where only 1 node of a cluster has full control of a resource at a time .. in this case the resource is the D: drive (that's the AX100 drive yes ?) And that the V: drive is internal to one of the cluster nodes (guessing it is another partition on the cluster nodes boot drive ?)
If you then tell VSS that it keeps track of changces to the D: drive by making using of space on V: the cluster software will (and should) tell you "No!" .. the reason being is that the V: drive can only ever be seen by the one cluster node .. the other cluster node could very well have a "V: drive" .. but it will be a completely difference V: drive... so in the even of a cluster failover where the second node takes control of D: .. how does it then get at the VSS data held on the V: drive of the other node ?!?!
A method that might work would be to create a new LUN on the AX100, presend it to BOTH cluster nodes as a drive letter (X: for example) and then make it a dependant resource in the MS Cluster .. where if D: is to be failed over, then X: must also fail over.
MarkNZ1
4 Posts
0
June 27th, 2007 16:00
MS Clusters works in a kind of active-passive arrangement .. where only 1 node of a cluster has full control of a resource at a time .. in this case the resource is the D: drive (that's the AX100 drive yes ?)
And that the V: drive is internal to one of the cluster nodes (guessing it is another partition on the cluster nodes boot drive ?)
If you then tell VSS that it keeps track of changces to the D: drive by making using of space on V: the cluster software will (and should) tell you "No!" .. the reason being is that the V: drive can only ever be seen by the one cluster node .. the other cluster node could very well have a "V: drive" .. but it will be a completely difference V: drive... so in the even of a cluster failover where the second node takes control of D: .. how does it then get at the VSS data held on the V: drive of the other node ?!?!
A method that might work would be to create a new LUN on the AX100, presend it to BOTH cluster nodes as a drive letter (X: for example) and then make it a dependant resource in the MS Cluster .. where if D: is to be failed over, then X: must also fail over.