[ntp:questions] Consider the Undisciplined Local Clock synchronized at start-up

Spoon devnull at localhost.com
Fri Mar 30 08:10:11 UTC 2007

Hello everyone,

I've configured one box to serve its local clock to another box.
(I want them to drift together.)

The server's configuration file is:

server              # LCL, local clock
fudge stratum 12   # increase stratum
logconfig all

When I start the "server" daemon, it takes a while (three minutes) for 
the local clock to be considered "synchronized".

09:59:14 : precision = 0.654 usec
09:59:14 : kernel time sync status 0040
09:59:14 : system event 'event_restart' (0x01) status 'sync_alarm, 
sync_unspec, 1 event, event_unspec' (0xc010)
09:59:15 : peer LOCAL(1) event 'event_reach' (0x84) status 'unreach, 
conf, 1 event, event_reach' (0x8014)
10:02:26 : system event 'event_peer/strat_chg' (0x04) status 
'sync_alarm, sync_local_proto, 2 events, event_restart' (0xc521)
10:02:26 : synchronized to LOCAL(1), stratum 12
10:02:26 : kernel time sync status change 0001
10:02:26 : system event 'event_sync_chg' (0x03) status 'leap_none, 
sync_local_proto, 3 events, event_peer/strat_chg' (0x534)
10:02:26 : system event 'event_peer/strat_chg' (0x04) status 'leap_none, 
sync_local_proto, 4 events, event_sync_chg' (0x543)

Thus, if I start the server and client together, the sender will refuse 
to reply to the client's iburst. (Please correct any misconception.)

I see that this issue has already been discussed in the past:
[Bug 767] New: wish iburst for local refclock

Is it possible to tell the server to consider his clock synchronized
as soon as the daemon starts? What is the appropriate keyword?


More information about the questions mailing list