[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: headless conf loses a cpu



On Sun, 2006-09-24 at 15:42 +0100, T. Horsnell wrote:
> >On Fri, 2006-09-22 at 15:54 +0100, T. Horsnell wrote:
> >> >On Fri, 2006-09-22 at 13:20 +0100, T. Horsnell wrote:
> >> >> >>What does /proc/cpuinfo say?
> >> >> >
> >> >
> >> >[ snip]
> >> >
> >> >> 
> >> >> Cheers,
> >> >> Terry.
> >> >> 
> >> >
> >> >Weird!
> >> >A. Can you access the machine BIOS from a serial port?
> >> 
> >> This will take me a while to find cable/connectors. In the meantime:
> >> 
> >> >B. Can you upload the output of /var/log/messages and /var/log/dmesg
> >> >with and w/o a VGA?
> >
> >Its 2am here and I'm a bit sleepy so I may be just guessing here... I'll
> >give your logs another read tomorrow... Maybe I'll find something
> >interesting.
> >
> >Anyways, of the top of my head it seems that you board/BIOS goes into
> >some weird safe mode, disabling the dual core, when the VGA card is
> >removed. I'd venture to guess that this board was never designed to
> >operate fully head-less.
> 
> The mobo documentation says it should operate w/o VGA
> 
> >
> >If this is indeed true I'd suggest you reinstall the VGA and leave it
> >disconnected. 
> >Nothing stops you from having a VGA card -and- having the
> >grub/kernel/console output redirected to a serial console. 
> 
> Agreed, and this is probably what I'll do. I dont even need serial
> console as ten of these are intended to be part of a compute farm.
> I just need to get hold of ten cheapo PCI low-profile VGA cards.
> Thanks for your time anyway.
> 
> Cheers,
> Terry.
> 

BTW, you might try to send a bug report to you mobo maker. It looks like
a BIOS problem to me.

- Gilboa


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]