[gpsd-users] Large ntp.conf time1 offset needed with GPSD shared memory driver

[Jon Brase]

Just a quick question: Is it typical for the data that GPSD hands to NTP  
through the shared memory driver to arrive with a ~500 ms offset from  
top-of-second as shown by PPS (where NTP is pulling the PPS data directly)  
or by external timeservers? The hardware is a Raspberry Pi running  
Raspbian Jessie with an Adafruit GPS breakout, with the following server  
configuration in /etc/ntp.conf:

------------------------------------------------------------------------------

#coarse time from GPSD
server 127.127.28.0 minpoll 4 maxpoll 4 prefer
fudge 127.127.28.0 time1 0.47 refid GPS

#PPS ref-clock for top-of-second
server 127.127.22.0 minpoll 4 maxpoll 4
fudge 172.127.22.0 refid PPS

# pool.ntp.org maps to about 1000 low-stratum NTP servers.  Your server  
will
# pick a different set every time it starts up.  Please consider joining  
the
# pool: <http://www.pool.ntp.org/join.html>
server 0.debian.pool.ntp.org iburst
server 1.debian.pool.ntp.org iburst
server 2.debian.pool.ntp.org iburst
server 3.debian.pool.ntp.org iburst

------------------------------------------------------------------------------

I found that I needed a time1 offset in the .45 to .5 range (currently  
.47, as shown above), to get the coarse GPS time to line up with the PPS /  
external timeserver time. Is this typical? I'm just wanting to make sure  
that it doesn't indicate some gross misconfiguration, as half a second  
seems like a rather large fudge. If everybody here thinks everything is  
fine, though, I think I will be ready to take leave of this list, as this  
is the last outstanding GPSD-specific issue I have with my GPS timeserver  
project.

On that note, what's the most appropriate forum for discussing general  
issues with setting up a GPS timeserver (for example, some of the output  
I'm getting from ntpq doesn't match what the references I used for setup  
lead me to expect, and I suspect misconfiguration or other problems with  
ntpd)?

--
Jon Brase