On Tue, Nov 02, 2010 at 01:29:14PM +0000, David L. Mills wrote: > This doesn't make any sense. Your data show the frequency set to 0.0 > and the a series of clock steps after that. If this is with your > simulator, try it with the real machine. I have tested with three > machines herewith and without kernel and with no such problems.
I have now such test running on real hw as I got that suggestion on #ntp too. But the result is the same, the frequency estimation is ~170ppm off, see the following log. It's running with "disable kernel" and the kernel frequency was left as ntpd using kernel discipline set it, so for the daemon loop the frequency offset would be close to 0. Again, this is with 0.1s initial offset, zero frequency offset, 64s polling interval and no drift file. Nov 2 12:53:08 localhost ntpd[8526]: ntpd [email protected] Mon Nov 1 12:20:56 UTC 2010 (1) Nov 2 12:53:08 localhost ntpd[8526]: proto: precision = 0.105 usec Nov 2 12:53:08 localhost ntpd[8526]: Listen and drop on 0 v4wildcard 0.0.0.0 UDP 123 Nov 2 12:53:08 localhost ntpd[8526]: Listen and drop on 1 v6wildcard :: UDP 123 Nov 2 12:53:08 localhost ntpd[8526]: Listen normally on 2 lo 127.0.0.1 UDP 123 Nov 2 12:53:08 localhost ntpd[8526]: Listen normally on 3 br0 10.34.24.25 UDP 123 ... Nov 2 12:53:08 localhost ntpd[8526]: Listen normally on 10 lo ::1 UDP 123 Nov 2 12:53:08 localhost ntpd[8526]: 10.34.32.125 8011 81 mobilize assoc 11921 Nov 2 12:53:08 localhost ntpd[8526]: 0.0.0.0 c016 06 restart Nov 2 12:53:08 localhost ntpd[8526]: 0.0.0.0 c012 02 freq_set ntpd 0.000 PPM Nov 2 12:53:08 localhost ntpd[8526]: 0.0.0.0 c011 01 freq_not_set Nov 2 12:53:09 localhost ntpd[8526]: 10.34.32.125 8024 84 reachable Nov 2 12:56:21 localhost ntpd[8526]: 10.34.32.125 963a 8a sys_peer Nov 2 12:56:21 localhost ntpd[8526]: 0.0.0.0 c614 04 freq_mode Nov 2 13:01:44 localhost ntpd[8526]: 0.0.0.0 0612 02 freq_set ntpd 172.556 PPM Nov 2 13:01:44 localhost ntpd[8526]: 0.0.0.0 0615 05 clock_sync Nov 2 13:20:18 localhost ntpd[8526]: 0.0.0.0 0613 03 spike_detect -0.138466 s Nov 2 13:29:04 localhost ntpd[8526]: 0.0.0.0 061c 0c clock_step -0.144407 s Nov 2 13:29:04 localhost ntpd[8526]: 0.0.0.0 0615 05 clock_sync Nov 2 13:29:05 localhost ntpd[8526]: 0.0.0.0 c618 08 no_sys_peer Nov 2 13:29:05 localhost ntpd[8526]: 10.34.32.125 8044 84 reachable Nov 2 13:35:47 localhost ntpd[8526]: 10.34.32.125 965a 8a sys_peer Nov 2 13:44:45 localhost ntpd[8526]: 0.0.0.0 0613 03 spike_detect -0.152547 s Nov 2 13:49:09 localhost ntpd[8526]: 0.0.0.0 061c 0c clock_step -0.177271 s Nov 2 13:49:09 localhost ntpd[8526]: 0.0.0.0 0615 05 clock_sync Nov 2 13:49:10 localhost ntpd[8526]: 0.0.0.0 c618 08 no_sys_peer Nov 2 13:49:10 localhost ntpd[8526]: 10.34.32.125 8064 84 reachable Nov 2 13:57:07 localhost ntpd[8526]: 10.34.32.125 967a 8a sys_peer Nov 2 14:09:12 localhost ntpd[8526]: 0.0.0.0 0613 03 spike_detect -0.130677 s Nov 2 14:15:45 localhost ntpd[8526]: 0.0.0.0 061c 0c clock_step -0.180840 s Nov 2 14:15:45 localhost ntpd[8526]: 0.0.0.0 0615 05 clock_sync Nov 2 14:15:46 localhost ntpd[8526]: 0.0.0.0 c618 08 no_sys_peer Nov 2 14:15:46 localhost ntpd[8526]: 10.34.32.125 8084 84 reachable Nov 2 14:21:22 localhost ntpd[8526]: 10.34.32.125 969a 8a sys_peer Nov 2 14:32:30 localhost ntpd[8526]: 0.0.0.0 0613 03 spike_detect -0.131304 s Nov 2 14:37:54 localhost ntpd[8526]: 0.0.0.0 061c 0c clock_step -0.155809 s Nov 2 14:37:54 localhost ntpd[8526]: 0.0.0.0 0615 05 clock_sync Nov 2 14:37:55 localhost ntpd[8526]: 0.0.0.0 c618 08 no_sys_peer Nov 2 14:37:55 localhost ntpd[8526]: 10.34.32.125 80a4 84 reachable Nov 2 14:49:16 localhost ntpd[8526]: 10.34.32.125 96ba 8a sys_peer -- Miroslav Lichvar _______________________________________________ questions mailing list [email protected] http://lists.ntp.org/listinfo/questions
