Bug or hardware failure

[Alvin]

On Thursday 15 April 2010 10:57:24 Alvin wrote:
> A week ago, a server of mine suddenly started to halt on random moments.
> Blank screen, no input. Drives and memory where fine. I needed frequent
> reboots to be able to finally start the machine (always that blank screen)
> Nothing in the logs hours before a sudden crash, nothing in /var/crash.
> 
> After a BIOS upgrade, the only message I got after a flash of grub was:
> PANIC: early exception 08 rip 246:10 error ffffffff810356e6 cr2 0
> (Older kernel gives same message.)
> 
> A replacement board showed the same error on boot, so unless both boards
> are faulty, a software error is more likely.
> 
> I had more success getting the system to boot with 'Intel Trusted
> Execution' disabled.
> 
> So, how would you debug a situation like this?

Well, the unexpected happened. It was a hardware error. Both boards (or CPU?) 
were faulty. I received a replacement board and CPU.