[ntp:questions] NTPD can take 10 hours to achieve stability
cswiger at mac.com
Tue Apr 19 00:49:23 UTC 2011
On Apr 18, 2011, at 5:42 PM, Mike S wrote:
> At 08:30 PM 4/18/2011, Chuck Swiger wrote...
>> Why would the local clock source "moving around" have any impact on a higher-stratum NTP time source running on some other machine?
> It doesn't, and that's not what was asked, said, or claimed.
Ah. So you're retracting these earlier remarks, then:
"Of course it is. Don't confuse the RTC used for power-off timekeeping (or the 127.127.1.0 "local" NTP driver) with the clock source used for timekeeping while operating. NTP doesn't keep time, it keeps the local clock source _in time_. NTP itself doesn't have an accuracy, one is concerned with how accurately it keeps the local clock in sync. If the local clock source is moving around, then NTP is constantly "chasing" it to try and keep it in sync. This has to have an effect on accuracy."
After all, with the caveat already noted about stratum-1 timesources, ntpd is sync'ing to timeserver(s) over the network and the accuracy of these remote timeservers is completely independent of the local clock.
More information about the questions