Re: NTPsec reports excessive jitter from GPSD/KPPS SHM

[Gary E. Miller]

Yo Nick!

On Tue, 10 Mar 2020 18:15:27 +0000
"Nick Burkitt" <address@hidden> wrote:

> I think it's time to stick a fork in this one.

Almost.  A few things still hanging.

You still have this near the top of your ntp.conf:

    # GPS Serial data reference (NTP0)
    refclock shm unit 0 refid GPS noselect

It should be after all your other time sources.

> ntpmon numbers seem much more reasonable now:
> remote  refid st t when poll reach delay   offset   jitter
> *SHM(2) .PPS.  0 l    2    8  377    0ns -10.37us  7.337us
>   SHM(0) .GPS.  0 l   60   64  377    0ns -60.73ms 6.4122ms

8 microseconds is still a tad high, but not bad.  That may settle with
time.

From your ntp.conf:

server 0.us.pool.ntp.org iburst

iburst is for laptops, that is counterproductive when you boot.  It
will make the ntpd first lock to the pool, instead of your PPS, which
is not what you want.

> I've uploaded a 1/2 day's ntpviz stats to 
> https://drive.google.com/open?id=1YXOxaQq94zurABpwtocW3gbpSDPSW5dO. 

Ugh.  Any way to view them as anything but raw html?

Also, you need at least 24 hours worth.  better yet, a week.

> Configuration details are in ntp.txt, graphs in ntpgraphs/. The drift 
> file was inaccessible for most of the period due to incorrect
> ownership, so it was only updated once.

It will take days to converge.

> Thank you for your help. I greatly appreciate the time and effort you 
> put into this discussion, and all the other discussions on the list.
> I can't imagine getting this kind of access to a principal developer
> for a commercial product.

Glad we could help.

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        address@hidden  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin

pgphANWTYFH5P.pgp
Description: OpenPGP digital signature