[ntp:questions] Re: Problem starting ntpd

Dale Worley worley at dragon.ariadne.com
Wed Aug 20 18:02:54 UTC 2003


I'm not expert in NTP, but I'd point out the following two items:

> The last line in /var/adm/ntp.log is:
> 
> 20 Aug 15:52:53 ntpd[12475]: parent died before we finished, exiting

When NTP starts up, it forks a subprocess.  The parent process "does
NTP", except that it is only in its default (not very useful)
configuration.  The child process reads the ntp.conf file, and passes
each line of ntp.conf to the parent as a configuration change.  The
child exits when it finishes with ntp.conf.  This system has the
advantage that some bits of ntp.conf processing can be quite slow
(e.g., host name lookups).

In this case, it appears that the parent is exiting before the child
is finished, and the child is complaining.  This is consistent with
the idea that the parent is terminating very quickly.

> Starting the daemon with 'ntpd -dddd' mostly, but not always, gives:
[...]
> Segmentation Fault (core dumped)

Something very bad is happening in NTP, because it should never seg
fault.  As you say, it probably has something to do with the refclock
driver.

Have you been able to recover a core file, or start ntp under GDB?
You might get a quick pointer to where the code fails.

Dale



More information about the questions mailing list