I'm referencing the NetApp Solution Guide, "Microsoft Exchange Server, SQL Server, SharePoint Server Mixed Workload on Microsoft Hyper-V and NetApp Fabric MetroCluster" and have some questions.
Based on my reading the solution is a four-node Microsoft Hyper-V multisite failover cluster with two nodes at each site running as an active-passive cluster (p. 6).
I'm assuming that the Hyper-V cluster is a geocluster (different subnets) and not a stretched VLAN (same subnet for both sites). Is that correct?
The solution relative to SQL 2008 calls for two SQL 2008 VMs (p.8) but it doesn't indicate that the SQL vms are clustered. I don't really get the SQL server setup. These are non-clustered SQL servers which seem to both be live at the same time and seem to be accessing the same data (e.g: "Separate aggregates are created for Exchange Server and SharePoint Server in site A and SQL Server and SharePoint are served from a single aggregate in siteB"). So both sites have live data and both sites have live SQL servers (which don't seem to be aware of each other). Also, since these SQL servers aren't clustered and therefore don't have application HA but only the HA provided by being vms on a HA Hyper-V server which is in a failover cluster how is the solution a site-resilient HA solution?
The document does mention that the VMs are hosted on NetApp iSCSI-based storage, which would be a requirement for clustering of Hyper-V guest VMs but, as I mentioned, I don't see any reference to the guest VMs being clustered.
I'd appreciate any clarification on the MetroCluster setup in general, SQL server HA as part of a MetroCluster and pointers to some in-depth MetroCluster documentation which I have not been able to locate on my own.