[ntp:questions] reach=0 and still synchronized

Krejci, Pavel pavel.krejci at siemens-enterprise.com
Tue May 25 18:41:54 UTC 2010


Hello,

when my own reference clock driver (for the /dev/rtc) stops to respond, the ntpd still shows that the clock is synchronized:
linux:~ # ntpq -pn
     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
*127.127.45.0    .RTC.            6 l  260   16    0    0.000   -0.074   0.059
The reach=0 and still the asterisk (*) is displayed at the beginning of line.

The ntpd shows:

/opt/ASwebmin/bin/ntpd_rtc: select() to /dev/rtc to wait for clock tick failed, errno=4: Interrupted system call.
refclock_transmit: at 293 127.127.45.0
addto_syslog: peer REFCLK(45,0) event 'event_unreach' (0x83) status 'unreach, conf, sel_sys.peer, 4 events, event_unreach' (0x8643)

peer REFCLK(45,0) event 'event_unreach' (0x83) status 'unreach, conf, sel_sys.peer, 4 events, event_unreach' (0x8643)
/opt/ASwebmin/bin/ntpd_rtc: select() to /dev/rtc to wait for clock tick failed, errno=4: Interrupted system call.
auth_agekeys: at 300 keys 1 expired 0
timer: interface update

but still the synchronized (asterisk) status is reported.
I have tried to call refclock_report(peer,CEVNT_TIMEOUT) in my poll message routine for the reference clock driver but nothing changed.
How could I force the ntpd to desynchronize the clock when the reference clock is no more reachable?

Thanks
Pavel Krejci




More information about the questions mailing list