Achim Gratz via devel writes:
> I can pull the info from my posts if you wish. But as a short summary,
> the server sends a "rate exceeded" KOD package (or the client interprets
> the server response in that way), hpoll gets set to 10 in response to
> that and the actual poll interval is at least 1024s (the longest I've
> seen was poll=13 or somewhat over two hours).
So it happened again yesterday, one of the boxes talking to the rasPi
1B+ got thrown off again:
raspberrypi3
remote refid st t when poll reach delay
offset jitter
===========================================================================================
oNMEA(0) .NavS. 0 l 10 16 377 0ns
-51ns 396ns
raspberrypi1.local 192.168.178.36 2 u 937 16 377 525.46us
-17.15us 100.55us
+raspberrypi2.local .NavS. 1 u 8 16 377 392.14us
26.550us 7.407us
-raspberrypi4.local .NavS. 1 u 7 16 377 320.36us
-3.946us 31.601us
raspberrypi5.local .NavS. 1 u 6 16 377 448.52us
76.654us 7.224us
+tinkerboard1.local .NavS. 1 u 5 16 377 363.35us
61.258us 5.788us
+tinkerboard2.local .NavS. 1 u 4 16 377 381.97us
27.260us 4.735us
ptbtime1.ptb.de .SHM. 1 u 43 64 377 23.151ms
2.7137ms 416.55us
ptbtime2.ptb.de .PTB. 1 u 61 64 377 25.300ms
2.4918ms 266.30us
ptbtime3.ptb.de .PTB. 1 u 39 64 377 24.459ms
3.1950ms 495.56us
-local 194.25.134.196 3 u 30 64 377 475.15us
-1.261ms 41.789us
A while later (after another poll) I logged into and ran ntpmon in
detail mode on the corresponding peer:
remote refid st t when poll reach delay offset jitter
oNMEA(0) .NavS. 0 l 9 16 377 0.0000 0.0001 0.0002
raspberrypi1.lo 192.168.178.33 2 u 234 16 377 0.6357 -0.0721 0.0986
+raspberrypi2.lo .NavS. 1 u 7 16 377 0.4002 0.0149 0.0093
+raspberrypi4.lo .NavS. 1 u 6 16 377 0.3755 0.0336 0.0041
raspberrypi5.lo .NavS. 1 u 5 16 377 0.3856 0.0357 0.0268
-tinkerboard1.lo .NavS. 1 u 4 16 377 0.3762 0.0604 0.0061
+tinkerboard2.lo .NavS. 1 u 3 16 377 0.3824 0.0293 0.0049
ptbtime1.ptb.de .SHM. 1 u 37 64 377 23.6188 2.9417 0.7565
ptbtime2.ptb.de .PTB. 1 u 53 64 377 25.3001 2.4918 0.2101
ptbtime3.ptb.de .PTB. 1 u 30 64 377 23.9516 3.0663 0.5310
-local 194.25.134.196 3 u 21 64 377 0.4704 -1.3027 0.0517
ntpd ntpsec-1.1.2+ 2019-01-13T16:01:41Z Updated: 2019-01-15T19:57:55 (8)
assoc=63109: conf, auth, reach, sel_reject, 5 events, rate_exceeded
dstadr=192.168.178.29:123 srcadr=192.168.178.28:123
reftime=2019-01-15T18:54:08.9Z leap=no-leap rootdelay=0.473
rec=2019-01-15T18:54:54.06Z stratum= 2 rootdisp=2.487
xmt=2019-01-15T18:54:54.06Z precision=-18 dispersion=24.378014
unreach=0 hmode=3 pmode=4 hpoll=10 ppoll=4 headway=7954 flash=4096 keyid=0
filtdelay = 0.67 0.88 0.64 0.53 0.66 0.65 0.69 0.77
filtoffset = 0.02 -0.20 -0.07 -0.02 0.06 -0.01 0.03 -0.18
filtdisp = 0.01 15.34 30.67 46.00 61.33 76.66 91.99 107.32
So again, a rate_exceeded event is logged and hpoll has moved up to 10.
Note that both ntpq and ntpmon show the poll interval still as 16s and
full reachability, which isn't true. Also, looking at the peer stats
you'll see that the effective poll interval actually changes once in a
while. 1024s is slightly above 17 minutes, but the distance betwee
successful pols that get recorded in peerstats is often larger than
that:
(one xtick in the plot is 15 minutes).
Regards,
Achim.
--
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+
Wavetables for the Terratec KOMPLEXER:
http://Synth.Stromeko.net/Downloads.html#KomplexerWaves
_______________________________________________
devel mailing list
devel@ntpsec.org
http://lists.ntpsec.org/mailman/listinfo/devel
