Re: [chrony-users] Chrony on Asus router - no reduction of offset

[Torsten Wolf <torsten.wolf@xxxxxxx>]

I didn't want to respond too quickly and observe the situation...

The current setup uses the local raspberry (with attached GPS receiver) 
as single source. This is the output from the server (public servers 
just shown for comparison)

chronyc -m sources sourcestats tracking
MS Name/IP address Stratum Poll Reach LastRx Last sample
===============================================================================
#? GPS                           0   4   377    14    +12ms[  +12ms] +/- 
100ms
#* PPS                           0   4   377    14   +723ns[ +809ns] +/- 
526ns
^? zeit.fu-berlin.de             1  10   377   841  -3489us[-3489us] +/- 
57ms
^? ptbtime1.ptb.de               1  10   377   28m  -3797us[-3794us] +/- 
9661us
^? ptbtime2.ptb.de               1  10   377   746  -5759us[-5759us] +/- 
12ms
^? ptbtime3.ptb.de               1  10   377   599  -2460us[-2461us] +/- 
9170us
^? ns.tu-berlin.de               2  10   377   125  -3807us[-3808us] +/- 
14ms
Name/IP Address NP  NR  Span  Frequency  Freq Skew  Offset  Std Dev
==============================================================================
GPS 6   4    82   -276.813   1024.470  +1861us  8124us
PPS 44  26   688     +0.000      0.002     +0ns   802ns
zeit.fu-berlin.de 31  18  516m     +0.034      0.030  -1611us   404us
ptbtime1.ptb.de 25  17  414m     +0.022      0.037  -2073us   368us
ptbtime2.ptb.de 19  10  310m     -0.014      0.082  -2628us   510us
ptbtime3.ptb.de 19   7  309m     +0.032      0.062  -2319us   334us
ns.tu-berlin.de 23  12  396m     +0.019      0.046  -2058us   417us
Reference ID    : 50505300 (PPS)
Stratum         : 1
Ref time (UTC)  : Wed Jun 14 15:54:52 2023
System time     : 0.000000008 seconds fast of NTP time
Last offset     : +0.000000087 seconds
RMS offset      : 0.000000120 seconds
Frequency       : 0.804 ppm fast
Residual freq   : +0.000 ppm
Skew            : 0.002 ppm
Root delay      : 0.000000001 seconds
Root dispersion : 0.000015001 seconds
Update interval : 16.0 seconds
Leap status     : Normal


Now to the Asus router (polling just the raspi and acting as server in 
the local network). After the responses I got here I changed the config. 
The following reflects the current setup.

server 192.168.1.105 minpoll -1 maxpoll 2 filter 15
server zeit.fu-berlin.de iburst noselect
server ptbtime1.ptb.de iburst noselect
server ptbtime2.ptb.de iburst noselect
server ptbtime3.ptb.de iburst noselect
server times.tubit.tu-berlin.de iburst noselect
driftfile /opt/var/lib/chrony/drift
dumponexit
dumpdir /opt/var/lib/chrony
pidfile /opt/var/run/chrony/chronyd.pid
ntsdumpdir /opt/var/lib/chrony
makestep 1.0 3
allow 192.168.0.0/16
deny 192.168.1.104
deny 192.168.1.105
broadcast 60 192.168.1.255
logchange 0.5
lock_all

chronyc -m sources sourcestats tracking
MS Name/IP address Stratum Poll Reach LastRx Last sample
===============================================================================
^* 192.168.1.105                 1  -1   377     2    +36us[  +41us] +/- 
350us
^? zeit.fu-berlin.de             1   6   377    27  -3778us[-3809us] +/- 
57ms
^? ptbtime1.ptb.de               1   6   377    32  -2285us[-2304us] +/- 
8006us
^? ptbtime2.ptb.de               1   6   377    21  -2048us[-2044us] +/- 
8463us
^? ptbtime3.ptb.de               1   6   377    24  -2564us[-2596us] +/- 
7919us
^? ns.tu-berlin.de               2   6   377    14  -3809us[-3798us] +/- 
15ms
Name/IP Address NP  NR  Span  Frequency  Freq Skew  Offset  Std Dev
==============================================================================
192.168.1.105 16   8   115     +0.013      0.814     +9ns    28us
zeit.fu-berlin.de 42  17   44m  -2675.180    171.496    -23ms   237ms
ptbtime1.ptb.de 24  12   24m  -2772.983    394.076   -146ms   229ms
ptbtime2.ptb.de 28  12   29m  -2740.241    334.810    -74ms   236ms
ptbtime3.ptb.de 42  17   44m  -2665.681    143.979  +7634us   227ms
ns.tu-berlin.de 33  16   34m  -2759.361    216.632   -191ms   204ms
Reference ID    : C0A80169 (192.168.1.105)
Stratum         : 2
Ref time (UTC)  : Wed Jun 14 15:58:57 2023
System time     : 0.000006924 seconds fast of NTP time
Last offset     : +0.000005338 seconds
RMS offset      : 0.070425719 seconds
Frequency       : 11.810 ppm slow
Residual freq   : +0.013 ppm
Skew            : 0.865 ppm
Root delay      : 0.000361663 seconds
Root dispersion : 0.000058637 seconds
Update interval : 7.6 seconds
Leap status     : Normal

The router runs a script which logs the offset (ms) and drift (ppm) 
every 10 minutes. For almost half a day the offset is fluctuating 
between +/- 0.05ms. This is much better than what I observed prior to my 
initial posting. I guess I would be happy but then there are one or two 
occations per day where the offset suddenly exceeds the logchange limit, 
e.g.,  to something like -0.7s which chrony then gradually corrects. 
Here is an example from today with offset and drift:

09:30:00    -0.027553 9.819
09:40:00    -0.004318 10.866
09:50:00    -0.005717 11.193
10:00:00    150.749 12.41
10:10:01    -50.4177 12.298
10:20:00    -43.7645 12.356
10:30:01    -17.488 12.256
10:40:00    -54.5296 12.498
10:50:00    -106.559 12.223
11:00:00    -80.029 12.407
11:10:00    -0.05282 12.508
11:20:01    -691.29 12.335
11:30:00    0.018295 12.594
11:40:01    -0.068731 12.71
11:50:00    0.007319 12.483
12:00:00    -0.087129 12.335
12:10:01    0.013139 12.451
12:20:00    -0.057343 12.031
12:30:00    -0.006875 12.203

I have no explanation how this is possible. I don't see anything 
suspicious in the raspi chrony offset data which I monitor via grafana. 
Could it be a hardware issue? Voltage fluctuations? I have no clue. How 
can I debug this?



Miroslav Lichvar schrieb am 12.06.23 um 09:21:
> On Sat, Jun 10, 2023 at 08:09:04PM +0200, Torsten Wolf wrote:
> > Name/IP Address            NP  NR  Span  Frequency  Freq Skew  Offset Std
> > Dev
> > ==============================================================================
> > 192.168.1.105              44  19   46m  -2691.479    161.086    -91ms 242ms
> > zeit.fu-berlin.de          48  21   50m  -2636.146    123.263    +17ms 218ms
> > ptbtime1.ptb.de            57  25   60m  -2632.208    101.466    +45ms 232ms
> > ptbtime2.ptb.de            54  23   57m  -2651.780    111.097    -18ms 234ms
> > ptbtime3.ptb.de            55  24   58m  -2601.458    110.065   +114ms 233ms
> > ns.tu-berlin.de            49  21   51m  -2600.763    126.308    +79ms 245ms
> The skew is too large for the frequency to be accepted with that
> maxupdateskew you have configured. This causes the clock to drift away
> between updates (sawtooth wave).
>
> There is something wrong with the system or hardware. Stddev of 240ms
> to a local server is too high. Is ping to the server better?