[time-nuts] Poor's man NTP

giuseppe at marullo.it giuseppe at marullo.it
Thu Dec 16 23:37:12 UTC 2021


Hi,

just wondering if a PI4 could be a suitable NTP server for a small lab (and
maybe with some other NTP servers for my company, about 2000 clients).

Main use is for correct timestamp on logs/computer time sync.

 

I setup the NTP using Adafruit Ultimate GPS shield(with battery), and the
GPS/GLONASS Antenna(cheap one, not a timing antenna). Antenna is on a roof
window with a small metal base, outside.

 

What is the accuracy I could expect from it?

 

According to ntpstat:

 

synchronised to UHF radio at stratum 1 

   time correct to within 2 ms

   polling server every 8 s

 

Accuracy is *only* within 2ms after several hours (>24h). I was expecting
under 1ms.

 

This is the output of "ntpq -crv -pn" :

 

ntpq -crv -pn

associd=0 status=0418 leap_none, sync_uhf_radio, 1 event, no_sys_peer,

version="ntpd 4.2.8p12 at 1.3728-o (1)", processor="armv7l",

system="Linux/5.10.63-v7l+", leap=00, stratum=1, precision=-20,

rootdelay=0.000, rootdisp=2.090, refid=GPS,

reftime=e56642d9.eca2e4f7  Thu, Dec 16 2021 23:57:29.924,

clock=e56642e0.a6b8351c  Thu, Dec 16 2021 23:57:36.651, peer=13425, tc=3,

mintc=3, offset=0.001227, frequency=-11.963, sys_jitter=0.000954,

clk_jitter=0.001, clk_wander=0.000, tai=37, leapsec=201701010000,

expire=202206280000

 

     remote           refid      st t when poll reach   delay   offset
jitter

============================================================================
==

o127.127.20.0    .GPS.            0 l    7    8  377    0.000    0.001
0.001

0.europe.ntp.or .POOL.          16 p    -   64    0    0.000    0.000
0.001

1.europe.ntp.or .POOL.          16 p    -   64    0    0.000    0.000
0.001

2.europe.ntp.or .POOL.          16 p    -   64    0    0.000    0.000
0.001

0.it.pool.ntp.o .POOL.          16 p    -   64    0    0.000    0.000
0.001

1.it.pool.ntp.o .POOL.          16 p    -   64    0    0.000    0.000
0.001

2.it.pool.ntp.o .POOL.          16 p    -   64    0    0.000    0.000
0.001

3.it.pool.ntp.o .POOL.          16 p    -   64    0    0.000    0.000
0.001

+193.204.114.232 .CTD.            1 u   14   32  377   28.914   -0.214
0.321

+185.19.184.35   193.204.114.233  2 u   55   64  377    8.074    0.094
0.314

-193.204.114.233 .CTD.            1 u   18   32  377   26.846   -1.077
0.198

*193.204.114.105 .CTD.            1 u    4   32  377   24.642   -0.320
0.366

 

BTW I don't know why reftime and clock time differs so much, and I don't
know why the difference vary, from some seconds to several seconds if check
even within a short time span.

Apparently the Adafruit module does have a RTC module and I enabled it, even
though there are no much info on it.

 

How to tell if 2ms is the real performance? Any recommendation?

Already disabled wifi interface and cpu scaling.

TIA

 

Giuseppe Marullo

IW2JWW - JN45RQ

 

Ps: this is the tutorial I used: Howto Raspberry 4 NTP stratum 1
(kiokoman.eu.org)
<https://www.kiokoman.eu.org/index.php/per-non-dimenticare/24-howto-raspberr
y-pi-4-headless-ntp-stratum-1-gps-pps>  (in Italian, sorry)

 





More information about the Time-nuts_lists.febo.com mailing list