[ntp:questions] Tighter regulation?

David Lord snews at lordynet.org
Sun May 26 10:20:39 UTC 2013


Mischanko, Edward T wrote:

....


> I don't believe the room temperature changes much at this computer's
>  location.
> 
> Now if your computer
>> heats up and cools down by 20C over an hour that could change the clock
>> by a few ms. And ntpd is slow to respond to changes in clock rate.
> [Mischanko, Edward T] 
> 
> If I do maxpoll 8, then I get regulation within 1 millisecond.  What other
>  details would you like?  I can post my loopstats in the future if that
>  would be helpful?
> 
>> What happens if you do maxpoll 6? Give us details, not vague statements.
>>
> [Mischanko, Edward T] 
> 
> I do not know how to analyze the peerstats file, but I can post it if that
>  would help?

My maxpoll settings are internet servers = 12, my public and lan
servers = 10, public and lan peers = 8.

Above is not ideal if ambient temperature or system temperatures
vary too quickly. December-January and June-July are best periods
for minimum temperature change but November-March I often lose my
GPS signal.


In the ntpd distributions I have is "summary.sh" which is a script
to do a summary then delete the peer, loop and clockstats files.


"summary.pps.20130516 gps/pps connected server on lan"

loopstats.20130516
loop 1350, 3+/-34.8, rms 4.9, freq -36.10+/-0.343, var 0.199

peerstats.20130516
        ident     cnt     mean     rms      max     delay     dist     disp
==========================================================================
127.127.20.2    1350    1.135   18.939   62.924    0.000    2.607    1.873
127.127.22.2    1350    0.001    0.005    0.038    0.000    0.930    0.930
81.187.61.69     310    0.123    0.261    1.057    0.839    6.286    3.056
90.155.53.93      74    0.683    0.283    0.721   16.277   31.016   10.816
81.187.61.74     316    0.512    0.726    2.085    1.530    7.029    3.310
81.187.61.65      66    0.154    0.859    2.502    0.940   22.667   10.039
81.187.61.78      68    0.046    0.500    1.770    0.688   22.090   11.560


"summary.ntp0.20130516 ntp0.lordynet.org.uk"

loopstats.20130516
loop 145, 135+/-1693.8, rms 699.0, freq 2.48+/-0.184, var 0.120

peerstats.20130516
        ident     cnt     mean     rms      max     delay     dist     disp
==========================================================================
130.159.196.117   69    0.011    0.720    1.914   26.757   34.487   11.098
81.187.61.65     154   -0.379    0.570    1.484    0.611   20.081    4.823
130.88.200.6      75    5.671    2.277    5.958   23.255   34.796   10.529
195.173.57.232    69    0.296    0.711    1.759   16.140   29.913   10.937
81.187.61.69     130   -0.279    0.708    1.697    0.424   19.692    5.287
143.210.16.201    73    0.237    0.619    1.812   24.244   33.428   10.498


David

> 
>> Also try plotting peerstats and the offset as a function of time.
>>



More information about the questions mailing list