ESX Cluster HBA Freezes

[Ron Burke]

We have 4 BL685 G5 servers in a C7000 chassis, using QLOGIC HBAs, Brocade 4/24 Powerpack SAN Switches connecting to an EVA 8100. ESX 3.5.

In January the 4 servers froze up, and we lost the whole environment. It was determined the cause was bad memory. Memory was replaced.

This week, all 4 servers froze up as if the HBAs stopped working on each server at the same time. One server had a pink screen when it was rebooted.

VMWare blames the HP HBA agents, and I have sent their response to HP. HP is still reviewing logs.

Has anyone ever seen all servers in a cluster freeze at the same time? Losing the entire environment, and causing loss of all services on it?

[Ron Burke]

Just updated to new firmware. All of our power supplies are being recalled (24 total).

http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c01519680

Don't know if this is the problem. Replacement utility and HP's program is good.

https://h30090.www3.hp.com/pcc-CPS/Adapter.aspx

I'm hoping this is it, still haven't heard from HP.