[ntp:questions] NTP vs chrony comparison (Was: oscillations in ntp clock synchronization)

David Woolley david at ex.djwhome.demon.co.uk.invalid
Sat Jan 26 13:14:15 UTC 2008


Richard B. Gilbert wrote:

> ISTR that time stamps on financial transactions are required to be 
> within two seconds of the correct time.  With NTP that standard is not 
> too difficult to meet.

In 2006, it turns out that it was 3 seconds 
<http://tf.nist.gov/general/pdf/2125.pdf>, which makes me wonder why so 
many people were concerned about proving traceability when using NTP, as 
NTP will have stopped accepting times when it thinks their maximum error 
is only 1 second.

It seems to me that an appropriate ear and keyboard method could pretty 
much guarantee 3 seconds.  It got the impression, from the questions, 
that the state of the art was much better than that.  cron, ntpdate, and 
a check for large corrections should do.  (In a non-airconditioned 
domestic environment 2 seconds a week with weekly corrections isn't 
unreasonable, once the frequency has been calibrated.)

I think the optimum solution for that requirement would probably be to 
simply capture the most recent NMEA time string of a GPS receiver, plus 
a simple timeout and a check for error sentences. (You can probably add 
two to three seconds to the GPS value to take full advantage of the 6 
seconds peak to peak error!)




More information about the questions mailing list