Re: slink server crashing
On Fri, Jun 25, 1999 at 03:55:57PM +1000, Ed Breen wrote:
>
> Hi,
>
> We are running a slink debian distribution with kernel 2.0.36 on
> a celeron 333 as our file server. Our SCSI adapter is an Adaptec 2940UW
> PCI card, with this configuration:
>
> 2940UW Adapter
> |
> Seagate Barracuda 9.1gig UW
> |
> Quantum Viking II 9.1gig UW
> |
> Sony AIT Tape Drive
> -TERMINATION.
>
> The Adaptec configuration program correctly recognises all the devices,
> and the system runs fine 99% of the time. However, when the system is
> under heavy disk usage, it sometimes suddenly cold reboots, eg: today
> while gunzipping a large .gz archive on the Barracuda drive the system
> rebooted. similar examples include copying large files from
> Windows/Samba clients to the server, and copying 3-4 gigabytes of files
> from the Barracuda to the Viking, using the unix 'cp' command.
>
> We have checked all the termination, and log files. Physical
> termination all appears to be correct, and there are no clues in the log
> files...the system simply cold reboots without leaving any error
> messages.
>
> Has anybody ever experienced these symptoms? Could it be due to a bug
> in the 2.0.36 kernel, or is it more likely an undetected hardware issue?
Bad termination should not be able to make the system reboot.
You might want to try the 2.0.37 kernel and put newer aic7xxx
drivers for the Adaptec card. You can find the drivers from
the url's mentioned in the readme file which is in the kernel
sources.
Long shot: lots of disk i/o -> lots of vibrations -> makes the processor
loose one of it's contacts for a second -> boom!.
Note: debian-user is better forum for this, so let's move this there.
--j
Reply to: