NTP Performance

ASSI Stromeko at nexgo.de
Thu Nov 21 06:55:19 UTC 2019

Richard Laager via devel writes:
> I now have a second stratum 1 server, with an independent setup. This
> allows me to compare the two. Why does ntp1 have this very specific
> repeating pattern of local clock offset? It's roughly +7 us, -5 us, +2
> us, -4 us and then repeats again, over and over.

Could be GPS constellation, could be thermal forcing or something
entirely else.  The question you'd need to answer is whether the
frequency drift is forcing the offset (likely thermal) or the other way
around (GPS constellation changes or something like that).  If you do an
autocorrelation or FFT on the data you should get a feel for what the
apparent period of the disturbance is and whether it stays stable over
time.  You probablby don't have enough tdata to resolve it yet, but if
it's GPS constellation it should have very slightly under 12h period.

> We can also see that in
> the histogram, which has three local peaks, as opposed to ntp2's which
> looks like a normal distribution.

> Here are the two NTPviz URLs:
> https://ntp1.wiktel.com/day/
> https://ntp2.wiktel.com/day/
> The server hardware is the same... same chassis, same motherboard, CPUs,
> etc. Both have their respective PPS connected via a motherboard serial
> port. Both are running Ubuntu 18.04 with NTPsec 1.1.7.

To me they look very different from a superficial look.  One swings
about 2ppm, the other 0.2ppm.  Unless that can be explained by very
different temperatures for each crystal, I'd conclude these two are not
the same hardware where it matters.  As I've said before, once you've
got around five servers you can start to usefully correlate such events
among them.  Just two, not so much.

> I could try fiddling around with the polling interval. Next steps might
> be to try raising the polling the interval to 4 and/or lowering it to 1.

It is generally inadvisable to use too short polling intervals unless
rthe clock you are disciplining is exceptionally unstable.

> As a second question, if you look at the ntp2 weekly graphs, there is a
> single huge transient. Any idea what that might be? I've seen these
> about once or twice a week since I put this in a couple weeks ago:
> https://ntp2.wiktel.com/week/

Going from my experience, that could indicate a suboptimal antenna
position.  That can result in the GPS solution drifting; for normal GPS
modules you'd see the position moving and the time drifting with it,
most often the altitude -- this is a timing module, so if it is in
timing mode the position is locked and only the time is drifting.  Once
it finally gets a better constellation and recognizes the error it
either "snaps" back to the correct solution or drops out completely
(I've had to cold-start my GPS modules a few times untiul I've improved

+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

DIY Stuff:

More information about the devel mailing list