[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-smp
Subject:    Re: System still crashes...
From:       Ben Dooks <ben () oaktree ! co ! uk>
Date:       1999-08-31 11:18:53
[Download RAW message or body]

> Ok, a while ago I emailed about a problem with an SMP system crashing
> under Linux. It was suggested that the problem was a memory leak in the
> 2.2.11 kernel. Well, now I've patched 2.2.11, and now completely
> reinstalled (the kernel) using 2.2.12, and I'm STILL getting the
> problem! What's more, for the first time, it crashed within a few hours
> of powering up, so it looks like it doesn't happen "after a day or two",
> but just "randomly, every few days".

I was experienceing the same problems with a BP6 and a pair of C400As (setup
at bottom) which got so bad that even fsck couldn't run to clean up the damaged
filesystems. I eventually took the machine to bits and swapped most of it, 
thinking first it waas either the mothebroard, network card or cpus./..
eventually i fouind that the memory seems to be a bit on the doddgy side
and replacing it with another (knoiwn good) DIMM has cured the problems.

The board has been under test for a day now and seems to be running fine
(albeit diskless) so it looks as if it is ok.

> So, any OTHER ideas?? Bung CPU? Peripheral clash? Other libraries
> causing instability? 

I'm ruinning debaian-2.1 on one box (similar abit bp6 config) and an
hand-upgraded slackeware 3.4 on the other both with 2.2.10-ac10 kernels#
	
> 've included below my original description of the problem, which is
> still apt, although the NumLock as described, it doesn't seem to freeze
> TOTALLY now, but still goes very unresponsive...
> 
> Trevor Phillips wrote:
> > 
> > Hi! I upgraded my PC at work to a Dual Celeron system on the day kernel 2.2.11
> > was released, and so I compiled it up with SMP support. Everything is fine
> > most of the time, but the system crashes once every day or two. The crashes
> > seem to be more of a grind to a halt, than a sudden freeze, although the
> > grinding IS rapid. First sign is the mouse freezes, then I usually toggle Num
> > Lock, and slowly the responsiveness of NumLock stops altogether, and I have to
> > hard-reset.
> > 
> > The Machine:
> >    ABit BP6 MB
> >    Dual Celeron 400 (NOT Overclocked)
> >    Matrox G400 16MB Video
> >    Adaptec AHA294X SCSI
> >    Digital DV21041 Tulip (D-Link Card) Network
> >    SoundBlaster AWE64
> >    128MB RAM

my two machines are based on:

	Abit BP6
	64Mb PC100 SDRAM
	Dual Celeron 400 (A) (not overclocked)

one has:
	6.4Gb EIDE (debian-2.1, win98, win2k rc1)
	RivaTNT (STB4400)
	3Com 3C905
	Yamaha based PCI soundcard

the other (now ok)
	No HD
	RTL8139 Fast Ethernet, tho i've had a DEC21140-AF in it as well
	Generic SVGA (ISA card)


> > 
> > The OS:
> >    Kernel 2.2.11 (no additional patches) configured as per SMP instructions
> > (RTC, etc...)
> >    Booted via Loadlin with mem=127M
> >    Debian 2.1 (Slink), with parts of Potato (libc6.1, etc...)
> >    XServer XFree86 SVGA 3.3.4 (installed Binary in place of 3.3.3.1 one)
> > 
> > Any ideas?? I *have* been load-testing this machine; I've been running two
> > SETI@Home's on it almost constantly, mainly to stress-test the machine, so
> > it's almost always running at 100(200)%. I dual-boot into Win98, and I haven't
> > had any problems under Win98 (other than usual Micro$oft problems), although
> > admittedly I spend 99% of my time under Linux...

You are probably not seeing what is casuing the fault if you reboot often
as its probably some random event that linux is catching...

> > 
> > I *really* hate this sort of problem; occurs rarely, and no easy explanation.
> > Being new the the Wide World of SMP, I'm not sure if this is an SMP
> > instability or not. And if it IS, is it system, or kernel related??
> 

I don't think it is, you may want to try going back to 2.2.10 (with ac's patch
set) which has been fine for me, although the pricipal machines running it
often get rebooted for the weekend DIY sessions atm.

-- 
Ben 

As you exit the plane, please make sure to gather all of your
belongings. Anything left behind will be distributed evenly among the
flight attendants. Please do not leave children or spouses.

----- End forwarded message -----

-- 
Ben 

As you exit the plane, please make sure to gather all of your
belongings. Anything left behind will be distributed evenly among the
flight attendants. Please do not leave children or spouses.
-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
To Unsubscribe: send "unsubscribe linux-smp" to majordomo@vger.rutgers.edu

[prev in list] [next in list] [prev in thread] [next in thread]