[ntp:questions] drift value very large and very unstable

Andy Helten andy.helten at dot21rts.com
Mon Mar 10 17:13:44 UTC 2008

Unruh wrote:
> Time should never move backward. "Steps" should be fast drifts. (10000PPM
> if necessary but not 1000000PPM.

I don't know that this is unconditionally true.  From
quoting Dr. Mills:
> 4. In the design of the nanokernel code that leaves here, the time as 
> seen by the application is not stepped backward unless the kernel clock 
> is stepped backward more than two seconds. So, if the kernel clock is 
> stepped backward more than the step threshold and less than two seconds, 
> the time as observed by an application would effectively stand still or 
> advance at a low rate for the interval required. There are other 
> features designed to avoid inconsistent reading of the kernel clock via 
> various means in and out of the kernel. These features have been 
> implemented in the stock Tru64 kernel for the Alpha, but so far as I 
> know have not been implemented in other kernels.

Maybe I misunderstand the above statement or maybe it doesn't apply
here, but I read it as fast drifts only happen conditionally (i.e. that
condition being a time jump less than 2 seconds but greater than
128ms).  This makes sense, doesn't it?  If "steps" were always fast
drifts, then NTP would still require days or hours to cover a large jump
in time.  Additionally, if time *never* steps backwards, we wouldn't be
having this conversation because I would have never needed 'tinker step
0' or '-x' and so I would have never encountered my "problem" and so I
would have never signed up to this mailing list and never sent in a
question with the subject "drift value very large and very unstable".


More information about the questions mailing list