RE: [chrony-dev] makestep command sometimes makes chrony stop reading its sources

["Hattink, Tjalling (FINT)" <T.Hattink@xxxxxxxx>]

>> I have to correct myself. Chrony was drifting 300 uSec for half an 
>> hour, not 300 nSec ! I have to say that the system clock on this 
>> system is really bad. On other systems I've seen it drifting for only

>> 100 uSec for 10 minutes.
>
>?? Not sure what you mean by "was drifting 300usec"? drifting usually
>refers to rate of change. Do you mean chrony time was offset by 300usec
>which then decayed to aroung 1usec?

Lets refrase this. What I saw is that the offset was slowly increasing
until it reached peak of 300 uSec. This is caused by the temperature
change of the system clock oscillator. After the temperature is settled
the offset slowly decreases again to its initial value.

>
>
>>
>> I use a separate pc that reads out the ntp time from chrony every 
>> second and compares it with its own PPS corrected time using a 
>> in-house developed application used for timekeeping our systems. It
is 
>> quite precise and it shows a std dev of 10 uSec for the ntp time for 
>> chrony.
>
>I suspect that the "read out time" makes for most of that.
>

Yes, it is jitter caused by the network and the interrupt handling of
the operating system.

--
Tjalling Hattink

---
To unsubscribe email chrony-dev-request@xxxxxxxxxxxxxxxxxxxx with "unsubscribe" in the subject.
For help email chrony-dev-request@xxxxxxxxxxxxxxxxxxxx with "help" in the subject.
Trouble?  Email listmaster@xxxxxxxxxxxxxxxxxxxx.