Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 19 Apr 2004 16:13:51 +0545
From:      Bikrant <bikrant_ml@wlink.com.np>
To:        freebsd-isp@freebsd.org
Cc:        Matthew Seaman <m.seaman@infracaninophile.co.uk>
Subject:   Re: PPPoed crashed!!
Message-ID:  <200404191613.51606.bikrant_ml@wlink.com.np>
In-Reply-To: <20040419085001.GC51644@happy-idiot-talk.infracaninophile.co.uk>
References:  <200404191156.29962.bikrant_ml@wlink.com.np> <20040419085001.GC51644@happy-idiot-talk.infracaninophile.co.uk>

next in thread | previous in thread | raw e-mail | index | archive | help
On Monday 19 April 2004 14:35, Matthew Seaman wrote:
> On Mon, Apr 19, 2004 at 11:56:29AM +0545, Bikrant Neupane wrote:
> > I'm running pppoed on freebsd-4.9. It is running smoothly. However it
> > crashes sometime with a pppoed.core file dumpped at the root directory.
> > gdb -c pppoed.core shows that it crashed with signal 11 segmentation
> > fault. I'm running the package as it came with the os. I have not
> > upgraded any thing except the kernel which I have customized for my
> > needs.
> >
> > Is this a bug or am i missing some parameters in the kernel?
>
> It's hard to say quite what the problem is with the small amount of
> information you've included.
>
> Certainly none of the software supplied with the OS should ever crash
> with a segmentation violation, and it's definitely a bug if that
> happens reproducibly.
>
> On the other hand, segmentation faults are most commonly seen as a
> symptom of faulty memory sticks.  Crashes due to bad memory will
> happen quite unpredictably, although they will tend to happen more
> when the machine is under load and there's a lot of memory IO going
> on.
It worked fine for almost 40 days without any problem. But after that pppoed 
crashed almost once in a day. I thought it might be memory issue. So I 
replaced the server with a new P4 Dell Poweredge server with HT support. I 
built everything from scratch, the kernel. But the daemon has crashed twice 
in last one week. 

At peak time the number of running tun devices is around 40.
These are the stats. i collected from mrtg.
Max running process: 126	avg 76
max Cpu utilization: 7%     	avg 4%
Max sys. load: 0.35          	avg 0.1
System Memory:		512 MB
Max mem utilization: 382 MB   avg:355 MB
Max Mbuf in use: 726		avg: 490

Traffic:
	Max: 2000kbps	Avg: 437kbps
I think resource utilization is quite low.


Here is the output of back trace
This GDB was configured as "i386-unknown-freebsd".
Core was generated by `pppoed'.
Program terminated with signal 11, Segmentation fault.
#0  0x280e461f in ?? ()
(gdb) bt
#0  0x280e461f in ?? ()
#1  0x280e48c1 in ?? ()
#2  0x280e46af in ?? ()
#3  0x280e48c1 in ?? ()
#4  0x280e4fd9 in ?? ()
#5  0x280e34dc in ?? ()
#6  0x280e3470 in ?? ()
#7  0x280dda10 in ?? ()
#8  0x280dd66d in ?? ()
#9  0x280c4cc9 in ?? ()
#10 0x280c4be5 in ?? ()
#11 0x8049640 in ?? ()
#12 0x804a22b in ?? ()
#13 0x8048dba in ?? ()
(gdb)


Thanking you,
Bikrant Neupane


> If the crash you're seeing happens repeatably and always occurs in the
> same place, then it must be a programming error.  In which case,
> please report the problem using send-pr(1).  Getting a backtrace from
> gdb will be helpful -- as will keeping hold of the core file you have
> to make available to a developer should one request it.
>
> 	Cheers,
>
> 	Matthew



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200404191613.51606.bikrant_ml>