Re: [gpsd-dev] Possible error report

[Sanjeev Gupta]

On Sat, Nov 9, 2013 at 10:34 AM, Gary E. Miller <address@hidden> wrote:

Can yu give me a quick summary of the current issue?  You have PPS at
ntpd but did not know where it comes from?

Summary, as I currently see it:

1. I have a cheap GPS device, serial, modified slightly for 1PPS, http://www.sureelectronics.net/goods.php?id=99
   
2. I am running debian test, updated often.  gpsd if from the debian archive, packaged by  Bernd Zeimetz
   
3. Server runs gpsd and ntpd, is part of the Pool project.  For now, to test Andy's scripts, shm0 is marked noselect
4. You can see the server at: http://ntpmon.dcs1.biz/gpsd-gmapv3.php
   
5. I use gpsmon to monitor
6. I believed that I had 1PPS working well, with the shm1 segment showing reachability of 377 in ntpq
7. Everyone was happy
   
8. Because I wanted a prettier display, I did a git pull, build, etc, of the 3.10-dev tree.  Did NOT do an scons install
   
9. Stopped gpsd.  Started gpsd from the build directory.  Started gpsmon from the build directory.
10. I do not see a PPS indication in gpsmon, nor in gpspipe output
11. Yet, ntpd is happily sync'ed to 127.127.28.1
    

So my question is: Why is ntpd happy?  Since gpsd is not getting a 1PPS signal, who is feeding one to ntpd?

 

I can log in and look.  You can encrypt an email message to me using the
key I sign messages with, or IM me using OTR.  I'm pretty aranoid about
passwords...

Gary, if you could send me your ssh pubkey, I will add you with root access.  Please feel free to break the box, I can rebuild.

--
Sanjeev Gupta
+65 98551208     http://www.linkedin.com/in/ghane