Tim wrote: > On Wed, 2005-10-26 at 19:25 -0400, Claude Jones wrote: > >>I'm not sure when it changed, but, I clearly remember that when ntp >>would occasionally fail, there would be a message that would say that >>it had, and then the machine would continue to boot. > > > I'm on dialup, so my network is only ever on the net sometimes, but > rarely ever rebooted. So I left NTPD on all the time. Unfortunately, I > notice the following problems: > > Trying to boot means a very long delay as NTP waits for ages before > giving up and letting the boot process continue without it. (Yes, I > think some sort of "start it and leave it in the background" technique > would be a good idea.) > > In the past NTP would handle losing the PPP connection. It'd start > doing its business again when the PPP connection was reconnected, > automatically. Now, it needs manually restarting. If I can't resolve > this directly, and get desperate enough, I'll probably end up starting > PPP with an intermediary script that starts PPP then restarts NTP. One quick thought here, which may or may not be relevant. I have ntpd on all the time and I reboot at least twice per day as I move my laptop from home to office and back. I have my eth0 set up to not activate on boot and I do this manually after logging in to GNOME. I also happen to have two profiles set up, one for home (with a fixed IP and DNS) and one for the office (using DHCP). Once I configured eth0 to not activate on boot, my boot time is fairly short, even though I have ntpd and firestarter attempting to start on boot. I get the [FAIL] messages for both, but these pass quickly. This has been the case since circa FC2, when somebody posted this approach someplace. HTH, Marc Schwartz
