ntploggps affects timing stats?

Paul Theodoropoulos paul at anastrophe.com
Thu Jun 13 00:09:15 UTC 2019


On 6/12/19 17:00, Gary E. Miller wrote:
> Yo Paul!
>
> On Wed, 12 Jun 2019 16:53:42 -0700
> Paul Theodoropoulos <paul at anastrophe.com> wrote:
>
>
> frequency offset similar, etc. Do the GPS logs factor into the other
> statistics generated, in some way? I can't imagine they would -
> jitter is jitter, the whole world 'round...right?
> Weird.  Maybe ntploggps is keeping your CPU from doing a microsleep
> or cutting the clock speed?  Maybe it keeps gpsd from getting swapped
> out or?
>
> RGDS
> GARY

Maybe on the microsleeps, yeah. I've noticed that ntploggps uses a "lot" 
of cpu time relative to ntplogtemp for example, so it does introduce a 
continuous load. I have my cpu pegged at 'performance', and using cpu 
partitioning to keep ntpd on core 2 and gpsd on core 2, so it shouldn't 
be moving around. But hey, it's an RPI, I've squeezed the living 
daylights out of performance tweaking it. I think the microsleeps is 
probably the "culprit", which I'm happy to have, now to start monitoring 
ntploggps so it'll get restarted if it dies again!

-- 
Paul Theodoropoulos
www.anastrophe.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ntpsec.org/pipermail/users/attachments/20190612/657b2cdb/attachment.htm>


More information about the users mailing list