We already tried that but nothing came up...

On Wed, 26 Dec 2007 09:01:09 +0000
Martin Johnson <[EMAIL PROTECTED]> wrote:

> I would set the console to the serial port, turn kernel trace on (if  
> possible - not my specialty) and log the kernel crash messages with a  
> second PC so that you can post them here
> 
> 
> On 24 Dec 2007, at 14:07, Gregory Cavelier - Materiel.Net wrote:
> 
> >
> > Hello,
> >
> > We are trying to run a net5501 with a lan1641 but we are facing a  
> > problem : The
> > kernel crashes. What's strange is that we can provoke that by  
> > running a script
> > which sends emails.
> >
> > The logs show nothing special, the only thing we can do to get it  
> > back is to
> > hard reboot it.
> >
> > We did the exact same test (same environment, same kernel) with a  
> > net4801 and
> > a lan1621 and the problem doesn't show up!
> >
> > So, hoping that someone could help us with that (and is brave enough  
> > to set the
> > same environment) here are all the details :
> >
> > The network topology is as shown in :
> > http://cleia.org/soekris/test_network.pdf
> >
> > Yes, I know, putting a switch here is not clean but hey, it's just a  
> > test  :)
> > (and the "Fake ISP Gateway" only has one ethernet card).
> >
> > the net5501 is running a debian etch with a stock 2.6.23.9 linux  
> > kernel
> > (configuration file at : http://cleia.org/soekris/.config)
> >
> > The whole distribution can be found
> > at : http://cleia.org/soekris/flash_soekris.tar.bz2
> >
> > This archive contains 2 directories : sdb1 and sdb3. They can be  
> > copied on a
> > 2Gb flash card (this is what we're using) using the following  
> > partition scheme :
> >
> >   Device Boot   Start      End     Blocks   Id  System       Mount  
> > point
> > /dev/sdb1   *        1       62     249952+  83  Linux        /boot
> > /dev/sdb2           63      186     499968   82  Linux swap
> > /dev/sdb3          187     1007    3310272   83  Linux        /
> >
> > The other 3 "fake" machines are running a debian etch linux  
> > distribution.
> >
> > To do our test we use a perl script
> > (http://cleia.org/soekris/test_mail.tar.gz )
> > which we run on the "Fake Mail Sender" to the "Fake SMTP  
> > Server" (running
> > Postfix). To use this script : ./test_smtp.pl --count=500000
> > We usually run two instances of this script to get the crash; it  
> > happens
> > between 15 minutes and 10 hours...
> >
> > Thanks for reading that far,
> >
> > Regards
> >
> > -- 
> >
> > Gregory Cavelier <[EMAIL PROTECTED]> - sysadmin
> > Domisys SAS - http://www.materiel.net
> > _______________________________________________
> > Soekris-tech mailing list
> > [email protected]
> > http://lists.soekris.com/mailman/listinfo/soekris-tech
> >
> 
> 


-- 

Gregory Cavelier <[EMAIL PROTECTED]> - sysadmin
Domisys SAS - http://www.materiel.net
_______________________________________________
Soekris-tech mailing list
[email protected]
http://lists.soekris.com/mailman/listinfo/soekris-tech

Reply via email to