I think that PeterKeuning isn't looking for 2 x Open-E in failover and 1 x ESX talking to Open-E thru MPIO.

He's more looking for 2 x Open-E in failover and 2 x ESX in cluster - no single point of failure !