[ntp:questions] 0.0.0.0 in log messages, ntp v4.2.6p5

Hari Mahadevan harikvpy at gmail.com
Fri Sep 26 06:43:35 UTC 2014


On Thursday, September 25, 2014 2:27:01 PM UTC+8, Hari Mahadevan wrote:
> On Thursday, September 25, 2014 2:23:13 PM UTC+8, Hari Mahadevan wrote:
> 
> > Does anybody know why the IP address reported in NTP log messages is always 0.0.0.0?
> 
> > 
> 
> > 
> 
> > 
> 
> > For example the message written while stepping the clock looks like this:
> 
> > 
> 
> > 
> 
> > 
> 
> > Sep 25 13:03:47 ntpd[1818]: 0.0.0.0 061c 0c clock_step +340.491291 s
> 
> > 
> 
> > 
> 
> > 
> 
> > I looked into ntpd source and it's always passing a NULL for report_event()'s second parameter where it should be passing the peer with which the local clock is being sync'ed. Interestingly the peer information is available, yet it's not being passed as second argument. Why?
> 
> > 
> 
> > 
> 
> > 
> 
> > TIA.
> 
> > 
> 
> > 
> 
> > 
> 
> > Hari
> 
> 
> 
> The source code I'm referring to is in ntpd/ntp_loopfilter.c, line number 386. report_event is defined in ntpd/ntp_control.c, line number 2929.

I modified the code replacing the NULL in report_event() call with peer local variable and it works fine. Now the clock_step log message does print the IP address of the peer that was the source of the clock sync.

Anyone knows why this parameter is set to NULL, not just in this one report_event() call, but in all report_event() calls in the loopfilter function? Specifically, I want to know if it's due to potential run-time errors that can bring down the daemon.

Thanks!



More information about the questions mailing list