I inherited an environment last week that has quite a few issues, so naturally I'm starting with the backups. We already run Veeam which is a big plus, but the vSphere cluster has about a half dozen Windows failover clusters on it, each cluster has it's own iSCSI LUN which is connected directly to the node VMs at the guest level which of course means that Veeam B&R can't touch data on that LUN as part of a normal VM backup.
Currently we use a rather terrible freeware application as a "secondary backup" which just copies files over SMB, no revisions or verification, nothing other than copy a file from here to there, since unbeknownst to my predecessor Veeam isn't touching the clustered volumes this is the only backup we have of the data in those clusters.
The way I see it I have two options for handling this with Veeam and they both suck:
1) I could use a File Copy job, this works fine & would be similar to what is in place now, but that job type is very limited and what really kills it for most data is the lack of revisioning.
2) I could treat each node as a physical box and use Endpoint/Agent to back it up, this would be a lot more robust than #1, however since the iSCSI volume follows the cluster role I suspect Endpoint will log a failure every time it backs up a non-active node which would make this a nightmare to monitor/manage.
Am I missing something or is there no good way to tackle this scenario with Veeam?