This receiver wasn't hooked to an NTP server, so I don't have direct correlation. However, the number of satellites tracked didn't noticeably drop, and the signal strength was still in a usable range, so I doubt there would have been much impact on timekeeping.
Also, my GPS-refclocked NTP server was running from a GPS-disciplined oscillator which is specifically designed to "ride out" GPS signal dropouts, so I don't think the NTP data from that unit would yield much info. John ---- Ryan Malayter wrote: > On Apr 5, 6:51 pm, [EMAIL PROTECTED] (John Ackermann N8UR) wrote: >> For what it's worth, the attached plot shows what I saw. This >> was the only significant signal strength anomaly in over one year of >> data capture. >> > > Did the drop in strength cause any anomalies with ntpd? That is, was > sync with PPS "lost" during the signal drop? Do you have any loopstats > graphs that correspond to the same time period? > > Very interesting stuff... I just realized that all of the stratum-1s I > was using for my organization have GPS as the source. I've > subsitituted a stratum-1 that uses ACTS instead to one of my local ntp > servers, but I am wondering if that is enough so handle a significant > GPS interruption. I have four local NTP servers serving two locations > with good connectivity, and each as at least one unique upstream > server. > > As others have mentioned, though, this may be somewhat pointless: an > event that distrupts GPS for a long period is likely to have so many > other ill effects that I won't care much about my company's > network ;-). In building our disaster recovery plans, we had a bunch > of scenarios to which the answer was "we cease operations", because no > employee could be expected to worry about the company in the event of > truly major disaster. > > > > > _______________________________________________ > questions mailing list > [email protected] > https://lists.ntp.isc.org/mailman/listinfo/questions > > _______________________________________________ questions mailing list [email protected] https://lists.ntp.isc.org/mailman/listinfo/questions
