MailScanner causing server to crash

[Bill Omer]

I do have a console on the machine, but the machine doesn't sit there
dead, it power-cycles it's self.  That makes me want to believe that
it's not a kernel panic.

I'm going to try the version 3 series of mailscanner and see what kind
of results I get with that.  Since it doesn't fork (as I understand it)
it may not create such a load to cause the machine to reboot.



On Sun, 2003-01-05 at 15:46, Richard Keech wrote:
> Bill,
>
> Run the system in runlevel 3 so you have a text console.
> This will better let you see what messages are generated
> that might otherwise be lost.
>
> You need to determine if it is experiencing a kernel panic.
> So the console messages at the time of the event are vital.
>
> I suspect the only involvement with mailscanner is the load
> mailscanner places on the system.
>
> regards
>
> On Mon, 2003-01-06 at 08:25, Bill Omer wrote:
> > I'm having a major problem with MailScanner (indirectly) and I am hoping
> > that someone here may be able to help me with it.  It seems that when
> > MailScanner (4.10) is running, the server tends to reboot.  However I
> > don't believe that it is a problem with MailScanner.  I say this because
> > the server is running on a RAID5 setup with a Mylex DAC960 controller
> > with Barracuda SCSI drives.
> >
> > I suspect that the machine is rebooting due to the slowness of the RAID
> > array.  I'm not exactly sure though, but I do feel that it is directly
> > related.
> >
> > I'm currently running sendmail with spamass-milter with procmail to
> > filter out and delete spam, but I would like to be able to filter
> > viruses again.  I have tried disabling Virus Scanning and Spam Checks,
> > with both disabled the machine still crashes.
> >
> > The crashes are random, sometimes the server will stay up and work
> > correctly for a few days, sometimes it can't last an hour.  The machine
> > its self is a quad Xeon 500MHz setup with a gig and a half of ram,
> > running RedHat 8.0.  There is nothing in the log files to point anywhere
> > since the machine is rebooting before it can write to syslog.
> >
> > Could anyone offer any insight on this problem?
> >
> > Regards,
> > Bill Omer
> >
> --
> G. Richard Keech        Chief Instructor / Senior Consultant
> Red Hat Asia-Pacific    rkeech at redhat.com, richard at keech.cx
> Melbourne Australia     http://people.redhat.com/rkeech
> +61-419-036-463 mobile  +61-3-9370-5611 fax
> Legal:  http://apac.redhat.com/disclaimer