[ntp:questions] Re: Philosophical question about strata

Wolfgang S. Rupprecht wolfgang+gnus20051014T092415 at dailyplanet.dontspam.wsrcc.com
Fri Oct 14 16:58:44 UTC 2005


brad at stop.mail-abuse.org (Brad Knowles) writes:
> At 6:30 AM +0000 2005-10-14, Wolfgang S. Rupprecht wrote:
>
>>  Question to the hard-core ntp jocks: Would anything break within the
>>  internal of ntp if the stratum of a reference clock were to be
>>  forcefully changed from the traditional stratum-1 for a correctly
>>  running reference to, say, stratum-5 for a gps with really crummy EPE
>>  that might be giving good time, but then again might not?
>
> 	Stratum 5 would not be an appropriate number for a machine
> whose refclock has wigged out.  Stratum 16 would be more appropriate
> in that case.

I was thinking more of the situation where the GPS signal wasn't
totally unusable but where the gps knows that it has a solution with
huge error bars.  In a situation like this, using the time from the
GPS would still be better than nothing, but just about any other
stratum 1,2,3 would be preferable.

For a concrete example, sitting at a picnic table under some redwoods
with a gps plugged in gives one a truly awful signal.  Unless one has
some connection to the net (say wireless or cellphone), the only
available time source is that gps.  Assigning it a stratum of 5 seems
like it would do the right thing in terms of making it an undesirable
time source, but one that could still be used in a pinch.

And as I found out the hard way, having a totally unsynced laptop when
doing real-time gps tracking causes all sorts of problems.  Once the
laptop has drifted off of real time by 20 seconds, the gps timestamps
were so far off of kernel time that the program thought the gps had
failed.  Syncing the kernel time to the gps time fixed that problem.

-wolfgang




More information about the questions mailing list