On 06/28/2014 09:40 PM, John Andersen wrote:
On 6/28/2014 5:25 PM, James Knott wrote:
not because it is more accurate Ummm... As I mentioned earlier, that's from the National Research Council, which provides Canada's primary time standard. If they're wrong then they've got a serious problem. Also, as I mentioned, my Android phone, running clocksync from that same server, agreed with the 2 NIST sources. So, if 3 agree and one doesn't, where do you think the
On 06/28/2014 07:46 PM, John Andersen wrote: problem is?
If their clock disagrees with every other clock then they have a problem. I don't care if they were commissioned by God.
Are you still stubbornly using "time.nrc.ca" instead of time4.nrc.ca like their own page says you should use? Have you read all the notices on their page? http://www.nrc-cnrc.gc.ca/eng/services/time/network_time.html
Please read that where it says "However a new time service for the old *TIME *protocol is now offered from a new server, *time4.nrc.ca*. Users are encouraged to switch to the NTP protocol described below, for a more accurate time service." Then further down: "To configure it to synchronize your clock to NRC time, you must set the NTP server to the following host name (do not use the IP address, as this may change): * time.nrc.ca * time.chu.nrc.ca" I am using the time.nrc.ca server they recommend for NTP. I am not using the old time protocol, so I don't use that time4 server. If I read that correctly, that time4 server is for the old time protocol, but they recommend NTP, which I am using at time.nrc.ca. Also, as I mentioned, Clocksync, on my Android phone, also using time.nrc.ca agrees with the NIST sources. So, I have one device using time.nrc.ca agreeing with both NIST sources and one that does not. It does not sound as those the server is the problem. BTW, one of their servers is time.chu.nrc.ca. CHU is a time and frequency station similar in function to WWV in the U.S.. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
On 2014-06-29 03:56, James Knott wrote:
I am using the time.nrc.ca server they recommend for NTP. I am not using the old time protocol, so I don't use that time4 server.
Yes, that's ok. But look: Telcontar:~ # ntpq --peers remote refid st t when poll reach delay offset jitter ============================================================================== LOCAL(0) .LOCL. 10 l 567 64 0 0.000 0.000 0.002 +AmonLanc.valino 176.31.53.99 3 u 34 128 377 0.272 0.031 0.874 -time.nrc.ca 132.246.11.232 2 u 35 128 377 177.638 -3.763 52.567 -masip.celingest 158.227.98.15 2 u 43 128 377 79.936 2.896 50.169 +81.184.154.182. 88.198.230.201 3 u 27 128 377 77.500 -0.120 51.337 84.77.40.132 193.67.79.202 2 u 90 128 67 89.250 1.588 16.269 #147.83.123.133 193.67.79.202 2 u 40 128 377 92.078 12.045 5.933 +eudyptula.init7 73.121.249.250 3 u 93 128 377 103.453 -0.663 24.582 +merzhin.deuza.n 145.238.203.14 2 u 93 128 377 76.960 0.182 1.736 -dns1-ha.uk.syra 131.188.3.222 2 u 92 128 377 90.297 -8.021 49.727 +dns3.masbytes.e 81.19.96.148 3 u 97 128 377 119.571 -0.073 14.982 *skarabrae.draxi 131.188.3.222 2 u 88 128 377 95.776 -0.046 33.751 -herbrand.noumic 192.93.2.20 2 u 96 128 377 85.991 4.629 66.550 #mail1.ugh.no 85.12.35.12 3 u 92 128 377 88.990 -5.037 25.797 Telcontar:~ # That's what I get after a while. Now notice the first char on each line. If you have the documentation installed, look at file:///usr/share/doc/packages/ntp-doc/ntpq.html. The field is named "tally" and is described in this other document: file:///usr/share/doc/packages/ntp-doc/decode.html#peer and it says: The Select Field displays the current selection status. (The T Field in the following table gives the corresponding tally codes used in the ntpq peers display.) The values are coded as follows: Code Message T Description 0 sel_reject discarded as not valid (TEST10-TEST13) 1 sel_falsetick x discarded by intersection algorithm 2 sel_excess . discarded by table overflow (not used) 3 sel_outlyer - discarded by the cluster algorithm 4 sel_candidate + included by the combine algorithm 5 sel_backup # backup (more than tos maxclock sources) 6 sel_sys.peer * system peer 7 sel_pps.peer o PPS peer (when the prefer peer is valid) Notice item "3", a "minus". Well, in my peer list above, time.nrc.ca gets a minus, it is discarded. Of course, it could be due to the network distance, in my case. In yours, you'll have to find out. -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" at Telcontar)
On 6/28/2014 7:49 PM, Carlos E. R. wrote:
Of course, it could be due to the network distance, in my case. In yours, you'll have to find out. http://www.eecis.udel.edu/~mills/ntp/html/cluster.html
That page describes what goes into the cluster algorithm, which is what eliminated that server. Not sure I understand it. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org
participants (3)
-
Carlos E. R. -
James Knott -
John M Andersen