Re: Strange system behavior

[Steven Bellovin <smb%cs.columbia.edu@localhost>]

On Sep 20, 2010, at 8:09 37PM, Paul Goyette wrote:

> On Mon, 20 Sep 2010, Perry E. Metzger wrote:
> 
>> On Mon, 20 Sep 2010 13:21:24 -0700 (PDT) Paul Goyette
>> <paul%whooppee.com@localhost> wrote:
>>> (Resending without most of the original attachments, since that
>>> seems to have exceeded a message-size limit!)
>>> 
>>> 
>>> I have a new machine in my farm, and it is exhibiting some very
>>> strange behavior.
>>> 
>>> Essentially, this machine does nothing more than 'build.sh release"
>>> several times daily for port-amd64.  (See [1] for test results!)
>>> 
>>> About half of the time, the build fails due to some host utility
>>> receiving a "segmentation fault", and almost always it fails on the
>>> exact same command and at the exact same place in the build!  But
>>> re-running the failed command interactively succeeds without any
>>> problem.
>> 
>> I'm betting you have bad memory. Try running memtest86 for a couple
>> of days on the box and see what happens.
> 
> Memory was my first suspicion.  I started with 2x2GB of Kingston.  Then I 
> added 2x2GB of Corsair memory for a total of 8GB.  The problem stayed, so I 
> removed the original Kingston memory leaving only the 2x2G Corsair. Doesn't 
> seem to make any difference - all three result in the same failure symptom.

BIOS settings, maybe?

On a laptop, I once had two DIMMs.  With either alone, the system was fine; 
with both installed, I had memory problems.  As best I can tell, it was a 
capacitance issue; both together drove the system slightly out of spec.  I 
replaced one of the DIMMs with a higher-quality (and higher-priced) brand and 
all was well.

I do agree that your description is unusual for classic memory problems.  
Still, if you can spare the time to run memtest, it's a good first step.

                --Steve Bellovin, http://www.cs.columbia.edu/~smb