ProLiant Servers (ML,DL,SL)
cancel
Showing results for 
Search instead for 
Did you mean: 

BL20P - "the case of the disappearing disk drives"

 
Alan Snider
Occasional Contributor

BL20P - "the case of the disappearing disk drives"

Problem: I have bl20p servers that seem to mysteriously lose their hard drives and then intermittently recover.

Background: Currently, we have 13 bl20p servers. All blades were deployed with Rapid Deployment using sysprep'ed images of both win2k server and "advanced" win2k server.

After weeks of operation, two servers exhibited a similar problem: on reboot, the servers display normal POST messages then fail with "attempting boot from hard drive (c:)".

On moving the hard drives from the failing blades to another blade server or an ML370 results in a normal, successfull boot.

What I've tried so far:
- ran Compaq (aka HP) diags- no problems reported
- updated Blade bios, Smart Array 5i firmware
- cleared nvram
- attempted boot from nt boot floppy
- with drives in a functional server, ran win2k recovery console; issued fixmbr & fixboot
- pxe boot seems to repeatedly boot with "need to boot to production mode but a prod partition does not exist".

The mystery: one of these servers seems to have completely recovered on its own from this condition. The other server seems to find its hard drives once every 10 reboots approximately.

I suspect the 5i controllers are the culprit. Any assistance would be appreciated. I'm tearing out my hair.