[ntp:questions] NTP settings for machine with irregular, short connections to the Net

Harlan Stenn stenn at ntp1.ntp.isc.org
Thu Sep 13 04:11:44 UTC 2007


> Harlan Stenn wrote:
> > Why don't you test how well the laptop works as a "time transportation"
> > mechanism first?
> > 
> > And they can make their own decision about how well the alternatives work.
> > 
> > If the laptop does it for them, great.
> > 
> > If not, they might be able to use a modem driver or a low-cost (~US$80) GPS
> > receiver.

> The problem here is that the laptop is powered off during transport and
> you are dependent on how well the TOY is able to keep time until it's
> powered back up. At that point you have absolutely no idea how far off
> the system's clock is from UTC and you can only assume that the contents
> of the drift file give you an idea how far off the clock frequency is.
> Of course if you are interested only approximately to the nearest 10-20
> seconds it may not be a big deal. It's better than no correction at all.

My point exactly.  Test it.  Set it up, turn it off, wait a "sufficient"
quantity of time, consider temperature changes during transport, fire it
up again, and see how far off it is.

H



More information about the questions mailing list