On Saturday 03 Mar 2012 15:55:39 walt wrote: > On 03/03/2012 05:55 AM, Mick wrote: > > I am sure that the CMOS battery is dead in this old box and this may be > > related to it, but the error persists even when I restart chronyd after I > > set the date manually: > > > > chronyd[3105]: Fatal error: settimeofday() failed > > chronyd[3424]: Could not settimeofday > > chronyd[3105]: Fatal error: settimeofday() failed > > chronyd[3424]: Fatal error: settimeofday() failed > > > > This is what syslog shows: > > > > Mar 3 13:47:12 lappy chronyd[3424]: chronyd version 1.26 starting > > Mar 3 13:47:12 lappy chronyd[3424]: Set system time, error in RTC = > > 1169296056129876.500000 > > Mar 3 13:47:12 lappy chronyd[3424]: Could not settimeofday > > Mar 3 13:47:12 lappy chronyd[3424]: Linux kernel major=3 minor=2 patch=1 > > Mar 3 13:47:12 lappy chronyd[3424]: hz=100 shift_hz=7 > > freq_scale=1.00000000 nominal_tick=10000 slew_delta_tick=833 > > max_tick_bias=1000 > > Mar 3 13:47:12 lappy chronyd[3424]: Could not open IPv6 NTP socket : > > Address family not supported by protocol > > Mar 3 13:47:12 lappy chronyd[3424]: Could not open IPv6 command socket : > > Address family not supported by protocol > > Mar 3 13:47:12 lappy chronyd[3424]: Frequency -36.574 +- 0.028 ppm read > > from /etc/chrony/chrony.drift > > Mar 3 13:47:21 lappy chronyd[3424]: System trim from RTC = > > 1169296066498122.500000 > > Mar 3 13:47:22 lappy chronyd[3424]: System's initial offset : > > 2147481497.991868 seconds fast of true (step) > > Mar 3 13:47:22 lappy chronyd[3424]: Fatal error : settimeofday() failed > > > > Should I worry about this? > > Maybe the hardware clock is too far off? Does hwclock -w change anything?
Yes it is - but I forgot to say, I usually set the date manually following start up, then I run hwclock --systohc and then I restart chronyd. Lately chronyd has started coming up with this settimeofday error. :-/ -- Regards, Mick
signature.asc
Description: This is a digitally signed message part.
