I had an interesting discussion recently with a number of my colleagues around the requirement to place additional uplinks (vmnics) on a virtual switch that are not used by iSCSI binding into the 'Unused' state. One would think that it might be useful to team these uplinks, placing some in Standby mode so that in the case of a failure on the active link, the iSCSI traffic could move to the standby uplink. But the requirement is that other vmnics must be put into the 'Unused' state and not teamed. Why?
This requirement prevents the VMkernel port from floating across uplinks in the case of a failure. The reason for this is that if the physical NIC loses connectivity, it should be treated as a storage path failure, not a network failure. We want the Pluggable Storage Architecture (PSA) in the VMkernel to handle this event and failover to an alternate path to stay connected to the storage.
This approach enables customers to consider storage resiliency based on multiple paths to the storage, rather than basing it on the number of networks available to a single storage path.
Get notification of these blogs postings and more VMware Storage information by following me on Twitter: @VMwareStorage
