Hi All,
We have 3 Hyper-V servers 2 of which are part of a cluster, 2 physical adaptors have been dedicated to iSCSI and LUNs are presented to the hosts for shared storage. Disk failover is controlled by the iSCSI MPIO.
We have a requirement for VM's to have direct iSCSI connection(s) to access the data on the disks, i think this is called a RAW connection (correct me if i'm wrong). To configure this my thoughts were to create two vSwitches and have each vSwitch connect to the physical iSCSI interface (all iSCSI vSwitchs will have the same name across all hosts).
With this is mind it raises questions:
- Is the correct way to do this?
- How do the VMs deal with disk failover events if the host is controlling the disk pathing via MPIO? My thoughts were to enable the 'NIC Teaming' option by ticking 'Enable this network adaptor to be part of a team in the guest operating system' within 'Advanced Features'. Would this allow the VMs to continue to work gracefully in the event of a failover when the hosts utilises MPIO for redundancy/load balancing?
I look forward for all of you professional guidance.