Re: System goes into complete hang

["D'Arcy J.M. Cain" <darcy%NetBSD.org@localhost>]

On Mon, 7 Apr 2008 12:40:59 -0400
"Greg A. Woods; Planix, Inc." <woods%planix.ca@localhost> wrote:
> 
> On 7-Apr-08, at 11:38 AM, D'Arcy J.M. Cain wrote:
> > Perhaps I wasn't clear.  The audio stutter is just the most obvious
> > symptom.  The entire system goes dead for a moment approx. every
> > 32.63157 seconds.  Playing music and tracking when it stops was how I
> > worked out that number but everything goes dead.  I can't even get  
> > caps
> > lock to work for that moment.  In fact, it is exactly like a mini
> > version of the problem that this thread is about.  I don't have to be
> > playing music to see the problem.  It is much deeper than that.
> 
> That's a pretty "approximate" number that you've expressed as a value  
> to the nearest 100,000'th of a second! :-)

It was a straight division based on about twenty minutes of observation
so I guess it could be plus or minus 1/40th sec.

> Does it go on like that for "hours" at a time, always as regular as a  
> clock?

It goes on 24/7.  I didn't track it that long but whenever I do it
seems to be pretty regular.  If I am typingg the keyboard goes dead for
a moment and often when it comes back it doubles the last character
like it did for the 'g' in "typing" in this sentence.

> How long is this "moment" that it's dead?

That's harder to measure but about a second or two.  Long enough to
checck the caps lock key to see if it works.

> Are you running an Xserver on the console display when this happens?

Yes.  Do you think I should try running raw console for a while?

> Is the system keeping time reasonably well, especially without having  
> ntpd running?

The system doesn't keep great time anyway.  I have ntpd runing every 15
minutes.  However, after it hangs a few times it doesn't seem any worse
than usual.

-- 
D'Arcy J.M. Cain <darcy%NetBSD.org@localhost>
http://www.NetBSD.org/