[ntp:questions] More Granularity in the US in the NTP Pool

Jason Rabel jason at extremeoverclocking.com
Sun Sep 9 23:45:42 UTC 2007


> If you really care, the best route is a hardware reference clock; e.g. a 
> GPS receiver or, if you get a solid signal from Fort Collins, CO, a WWV 
> or WWVB receiver.  That makes you a stratum 1 server!  A GPS receiver 
> with a PPS signal should be accurate to within 100ns or better.  After 
> your computer mangles the signal, it should be accurate to 100us or
better.

Well, but the road travels both ways. A person using the pool would want the
best time possible (obviously). But also a person donating to the pool would
like to know the bandwidth being used is being put to good use (and not
merely discarded data).




More information about the questions mailing list