[ntp:questions] ntpd -x option

Unruh unruh-spam at physics.ubc.ca
Fri Mar 13 21:48:37 UTC 2009


David Woolley <david at ex.djwhome.demon.co.uk.invalid> writes:

>Matthias Fuchs wrote:

>> 
>> But now, when I start ntpd with the -x option and an offset of
>> about 30s before ntpd is started, I get error messages like this
>> 
>>         frequency error 557 PPM exceeds tolerance 500 PPM
>> 

>I should also have pointed out that the 500ppm slew rate limit is 
>imposed by *n*x kernels, not by ntpd.  In the case of -x, which forced 

No it is not. the kernel limit is 1 part per 10, by altering the tick rate.
There are two adjustment parameters, the tickrate and the frequency. ntp
uses only the frequency adjustment which is limited to 500PPM, and does not
adjust the tick rate.

At least that is how I read
man adjtimex

>the user space discipline, slews are implemented by periodically telling 
>the kernel to slew by a certain amount of time (not a rate).  The kernel 
>does this by changing the amount it adds to the high resolution software 
>clock for every tick until that correction is complete.  The amount of 
>the change is generally fixed at +/-500ppm (although it is less on some 
>older systems).

>This worked where the clock could only be read to a resolution of one 
>tick. However, I would want to check the kernel sourced code before I 
>would be prepared to say that this produces strictly monotonic time with 
>the tick interpolation schemes used on modern systems.




More information about the questions mailing list