Hi, On Mon, 24 Jan 2005 13:11:55 +0200 Silviu Marin-Caea <.> wrote:
Tip: to avoid running xntpd all the time on machines that don't need a very precise time synchronization, I have this script:
/etc/cron.weekly/synctime
ntpdate ntp.pick.a.server.close.to.you hwclock --systohc --utc
I made my own version of the above idea you kindly provided us. I wrote this, while sitting on a machine having a medium-frequency dial-up and quite rare LAN connection, sowishing to update my time 100% manually:
#!/bin/sh
/etc/init.d/xntpd start date ntpdate xxx hwclock --systohc --utc date /etc/init.d/xntpd stop
(Of course xxx is actually a correct domain-name, where the time-service is constantly running by my provider.) After attempting to run the script, I usually get this error-message I couldn't decypher: khazad-dum:/home/oscar # synctime Starting network time protocol daemon (NTPD) done Tue Jan 25 09:29:38 CET 2005 25 Jan 09:29:39 ntpdate[6373]: the NTP socket is in use, exiting Tue Jan 25 09:29:41 CET 2005 Shutting network time protocol daemon (NTPD) done khazad-dum:/home/oscar Do you have an idea, what do these few lines exactly mean, and how to avoid this kind of error? Thanks is advance, Pelibali