[ntp:questions] SNTP server + ntpd 4.2.4 client

Unruh unruh-spam at physics.ubc.ca
Mon Mar 17 18:00:59 UTC 2008


"Richard B. Gilbert" <rgilbert88 at comcast.net> writes:

>> 
>> It you attach a gps PPS receiver to one of your boxes (the server) and you
>> use a reasonable client then yes you can expect much better than 100us
>> accuracy on your net-- assuming it is not overloaded and the machines are
>> not overloaded with disk activity.
>>  
>> 
>> 
>> 
>>>Regards.
>> 

>If you REALLY want to determine the accuracy of NTP, get yourself a 
>Cesium Clock, have it calibrated by your national standdards lab, and 
>then measure your error.

??? Just get yourself a gpsPPS receiver. They cost about $60 ( vs thousands
for a cesium clock) and are more accurate than your ability to read the
system clock. Ie, anthing more accurate than about 1usec is completely
wasted on almost any computer around, since the random latencies in reading
the clock and servicing the interrupt are the dominant source of noise. 



>If you think that's too much trouble, you're probably right!!

Yes, especially since the additional accuracy over a $60 solution is
completely wasted.


Note to the OP. Apparently you are evaluating some time source. It probably
costs $2000 or more. That money is completely wasted. YOu do not need it to get the
accuracy you can use. Unless your system is under a mountain or in a
basement with no access to the sky, a gps PPS receiver will work just as
well as your expensive time source. 






More information about the questions mailing list