Hi everyone, reasonably new to VMWare and Open E on this sort of scale in practise but I understand a lot of the concepts.

What I am proposing to do:
Run 2x physical servers with Open E on them
I will also posess an NFS Failover license
So both devices are redundant

Run 2x physical servers with vSphere ESXi 5 installed to pool resources (eventually I would like to implement HA and vMotion once I prove that this will work the way I intend)
Each physical server will have two teamed gigabit connections to two different switches most likely a Cisco SGE20026 for network redundancy.

Setup (for example) 5 virtual servers running server 2008 R2, one virtual server running CentOS for Postfix, one virtual server running Zabbix. However, the actual OS is installed and all data is run from the Open E devices.

I am planning on using BCM5709 quad port NIC PCIe NIC cards as they are reasonably cheap and this system is just a proof of concept for something much larger. But I read this thread http://communities.vmware.com/thread...rt=60&tstart=0 suggesting that BCM5709's do not support iSCSI boot.

The ultimate goal is rundundant storage, redudant switching and redundant processing. Am I on the right track with what I am doing? Or can someone suggest a better implementation.

See link for the plan as it sits in my head. Please excuse crude nature of the diagram.

http://communities.vmware.com/servle...8028/diag1.png