[ntp:questions] NTPD can take 10 hours to achieve stability

David Lord snews at lordynet.org
Tue Apr 19 23:01:58 UTC 2011

unruh wrote:
> On 2011-04-19, David Lord <snews at lordynet.org> wrote:
>> David J Taylor wrote:
>>> "unruh" <unruh at wormhole.physics.ubc.ca> wrote in message 
>>> news:slrniqpa40.ppu.unruh at wormhole.physics.ubc.ca...
>>> []
>>>> Well, I think someone other than the current maintainers will have to
>>>> port it to windows. Since windows timekeeping is not the worlds best
>>>> anyway, it is probably true that the extra accuracy of chrony is
>>>> unnecessary. It does have a command line option "like ntpq -p"
>>>> provided by chronyc (depending on what you mean by "like").
>>>> What MRTG is I do not know.
>>>> If you are happy with ntpd, by allmeans stay with it.
>>> A pity that chrony will not be offered for Windows, at least for tests 
>>> to see whether it lives up to its claims.  There are times when a more 
>>> rapid convergence would be welcome, such as the reboot of PC Molde 
>>> around 13:30 yesterday:
>>>  http://www.satsignal.eu/mrtg/molde_ntp-b.html
>>> MRTG is a standard logging tool for network I/O which uses SNMP to 
>>> produce the graphs I have quoted here many times for network throughput 
>>> and timekeeping:
>>>  http://www.satsignal.eu/mrtg/performance_network.php
>>>  http://www.satsignal.eu/mrtg/performance_ntp.php
>>> I've written how to extend MRTG to monitor NTP timekeeping, and various 
>>> other parameters such as disk space and temperature here:
>>>  http://www.satsignal.eu/ntp/NTPandMRTG.html
>>>  http://www.satsignal.eu/mrtg/performance_howto.php
>>>  From chronyc I would need to be able to use a simple Perl script to 
>>> extract the numbers to be plotted - such as the Offset in the graphs 
>>> above.  An easy job if the format is standardised and machine readable.
>> I based my script on yours for ntp but changed from using a
>> linear axis to log. I first used chrony on NetBSD when on
>> dialup with demon and later for a while when on broadband.
>> When I first compared chrony with ntpd there was no contest
>> but more recent experiments with chrony had periods of
>> severe instability much worse than ntpd.
> More recent means what? What version of chrony? (there was a bug found
> in the past two weeks that was introduces a few months ago which did
> result in instability) 

More recent than  November 2009.
Probably between Dec 2009 and Jan 2010 and p4x2400c.

chrony 1.23

>> P4X2666 with chrony
>> <http://www.lordynet.org.uk/mrtg/stats/>
> Not at all sure what I am supposed to see. I have no idea what the graph
> axes represent?  What is 1.1k 

X-axis is in hours
Y-axis is offset in us so 1.1k = 1100us

Next experiment is to couple up the GPS again but if you
click on the ME6000 in the current stats the yearly graph
for Sep/Oct/Nov was when I last had the gps connected.


> Note that you can see my offsets with chrony on my web page. The network
> problems really do cause difficulty, but the offsets in general are very
> small -- 10usec over the network. 
>> Current stats all with ntpd.
>> <http://www.lordynet.org.uk/mrtg/stats/ntp/>
>> David

More information about the questions mailing list