system hangup - I'm lost

[cpghost]

On Thu, Oct 02, 2008 at 06:51:06PM +0200, Oliver Lehmann wrote:
> > today I'd a crash again - I was not able to get a crash dump (thought a
> > "panic" at the end of the kdb would do it but didn't - should have called
> > dumpon before ;)) - so here now the information I was able to retrieve:
> > 
> > Ok, what I've got so far is wrinting stuff out to the console when the
> > system hangs up:
> > 
> > swap_pager: indefinite wait buffer: bufobj: 0, blkno: 2, size: 4096
> > swap_pager: indefinite wait buffer: bufobj: 0, blkno: 2, size: 4096
> > swap_pager: indefinite wait buffer: bufobj: 0, blkno: 2, size: 4096
> > swap_pager: indefinite wait buffer: bufobj: 0, blkno: 2, size: 4096
> > ...
> > 
> > and now the debugger stuff:
> >
> > [snipped]
> >
> 
> So.. no idea? anyone?

If it's PATA, check the cabling, then check it again, and just to
make sure, replace the cable even if the system used to work flawlessly
in the past. I've had this on a few servers, but replacing the cables
always fixed the problem for me.

Oh, btw, you can reproduce this exact behavior on diskless workstations
with an NFS-mounted swap.

IIRC, it even happened on VERY slow hardware with GBDE or GELI-encrypted
swap partitions; but I'm not 100% sure it was due to slowness (it could
have been a bad cabling issue as well).

> -- 
>  Oliver Lehmann
>   http://www.pofo.de/
>   http://wishlist.ans-netz.de/

-cpghost.

-- 
Cordula's Web. http://www.cordula.ws/