Pre-Phase: tried using Openfiler on desktops as servers, worked pretty well.
TestPhase: Tried OpenFiler on 2 HP DL 185 G5's with a HP Smartarry P410 with 512MB BBW: Got "false" HD failures that would require a reboot and scsi abort tasks

Phase 1:
Switched to Open-e. Was pretty stable.
Switched from file_io to Block_io. Got HD's getting marked bad again.
Updated to newer Open-e Version, still got HD's getting marked bad, but would come up good on reboot.

Resolved by updating firmware on HP SmartArray P410 and on Sata HD's. This is a known issue with certain firmware version of P410 and Sata HD's.

Phase 2:
About 7 Months ago, Vsphere locked up, called vmware support. Problem traced to Open-e Lun.
Turned off all vms and hosts (some from Vmsphere and some through ssh to console of esxi). Rebooted and resync arrays and was resolved.
-failover would not work.
-All ip's from open-e got added as paths. Possible failover would not finish because non-virutal ip's were active i/o

Phase 3:
about a month ago had same problem.
*Took out dynamic targets for iscsi out of Vmware
* changed ping node list to only switches that should up 100% of the time.

Phase 4:
A few days ago had same problem and had to shutdown all vms.
* Open-e support noticed I had NIC bonding with Nics using two different drivers
* I have changed the bonding so nics are from same driver/device

Question: What other things should I look for to prevent the lun lockup? Failover would not work in these instances.
What other recovery actions should I take?
(if this happens next time I will probably just Physically failover by disconnecting all Network cables from primary).

Thanks