[ntp:questions] Re: NTP Stats within norm?

david at djwhome.demon.co.uk david at djwhome.demon.co.uk
Fri May 26 20:46:10 UTC 2006


In article <4476fee8$1_3 at ns2.industrial-logic.com>,
"Scott L. Glidden <test at test.com>" wrote:

> Yes, I am familiar with rules. This is a primary NTP server for 8 other 
> servers. I was using a NON PPS gps receiver and found the internet servers 

That's a definite contra-indication for using stratum 1 servers!  I'd 
guess the threshold is somewhere between 1,000 and 10,000 downstream 
systems.  Actually, many of the stratum one servers you were using
are the "obvious" ones, which tend to be overloaded and likely to 
perform worse than stratum 2 ones.

> worked better (Much less jitter).... Currently working on locating a PPS 

Specifically you want a time transfer receiver, although I suspect all
PPS ones are.

> device in the mean time... Was thinking about trying a bluetooth GPS 
> receiver to see if it worked any better than the RS-232 connected GPS box...

Most unlikely.  I think they are all navigation receivers and I'd expect
more complex protocols to add additional uncertainty.

> Bogus address (which one)

test.com. It appears to belong to Network Solutions and is presumably
reserved for test data.  If you want to use an invalid address live
you should use the .invalid. domain.  However, you will find all the
regulars use replyable addresses.




More information about the questions mailing list