[ntp:questions] GPS receivers were severely disrupted on 06 Dec 2006

Richard B. gilbert rgilbert88 at comcast.net
Thu Apr 5 21:44:32 UTC 2007

Ryan Malayter wrote:
> On Apr 5, 9:35 am, j... at specsol.spam.sux.com wrote:
>>Computer clocks drifting is going to be the least of the worries
>>if there is ever an extended and wide spread GPS outage.
> I would imagine there are some truly critical applications out there
> that rely on accurate time from NTP. And any critical application that
> relies on GPS should have an adequate backup in case of GPS failure.
> Airplanes/ships have intertial and radio navigation aids to backup
> GPS. Car owners should have (or can easily acquire) paper maps to back
> up their GPS. Surveyors have old-style mechanical devices to back up
> GPS.
> So GPS-disciplined stratum-1 NTP servers most likely have a failover
> arrangement with a non-GPS time soruce (such as USNO, NIST, WWV(B),
> DFC77, whatever) as a backup, right? But it appears at least some
> don't: ntpq reports that my organization's ISP has stratum-1 servers
> with GPS sources only. So they would, I think, start reporting
> 'unsynchronized' to clients in the event of a long GPS outage.

Right!  If you, or your organization, require accurate time no matter 
what, then you, or your organization, need to arrange some sort of 
backup.  This might be an atomic clock of your very own (Cesium or 
Rubidium), or an extremely accurate quartz oscillator.  Your choice will 
depend on your budget and your need for accuracy.  You may find that 
there IS no good solution within the constraints of your needs and your 
budget.  A great deal would depend on the duration of a GPS outage. 
Five or ten minutes would be easily survivable for most.  The longer the 
outage, the greater the cost.

More information about the questions mailing list