[ntp:questions] Re: Question about min/max poll and an interesting plot showing maxpoll's effect
Richard B. Gilbert
rgilbert88 at comcast.net
Thu Apr 20 22:02:08 UTC 2006
John Ackermann N8UR wrote:
> I am running a stratum 2 server that syncs to my pile o' stratum 1s. In
> addition, just in case my lab explodes, I have a couple of external
> clocks. I've set maxpoll to 6 on the server lines for the internal
> machines, to hold the poll interval at 64 seconds.
> However, doing so seems to cause the polling interval for the other
> servers -- even though they don't specify a shorter maxpoll than the
> default, they continue to poll at 64 seconds. Is that normal?
> By the way -- if you want to see the effect of more tightly controlling
> ntpd through shorter polling, take a look at
> http://www.febo.com/time-freq/ntp/stats/clients. These plots update
> every hour, so may look different by the time you see them, but at least
> for now they show a dramatic improvement in stability when switching
> from normal polling (which ran out to 1024 seconds) down to maxpoll 6.
> The level 2 server is holding within about +/- 4 microseconds to about a
> 4 microsecond offset from the server (a stratum 1) that is monitoring
> it. Looking at the 7 day plot shows the dramatic change when I made the
> maxpoll adjustment. (Of course "dramatic" is a relative word -- even
> polling at 1024 seconds, the server stayed within about a millisecond).
> questions mailing list
> questions at lists.ntp.isc.org
Ntpd is supposed to set it's own polling interval; something to do with
the "Allen Variance". My feeble understanding translates into English
as "short poll intervals allow ntpd to correct large errors quickly and
long poll intervals allow ntpd to correct small errors accurately".
Anyway, I'm a little surprised that shorter is better. I have two
machines as clients of my stratum 1 (GPS) server. All three are Sun
workstations; the server is an Ultra 10 440 MHz, one of the clients is
also an Ultra 10 440 MHz the other is an Ultra 5 360 MHz. All have been
running for several days. The Ultra 5 is polling at 64 second intervals
with an estimated error of 11056 us while the Ultra 10 is polling at 256
second intervals with an estimated error of 500 us.
The proper poll interval appears to be highly dependent on the quality
of the local clock.
More information about the questions