Bugzilla – Bug 140128
ntp does not start correctly at boot
Last modified: 2006-01-16 14:51:48 UTC
The ntpd process does not start correctly on boot. This happens on all of my 10.0 systems. The process is started, but "ntp -q" only shows the "LOCAL" clock and none of the configured servers. In addition, manually running "rcntp restart" needs to be done several times to get the daemon properly initialized.
(In reply to comment #0) > The ntpd process does not start correctly on boot. This happens on all of my > 10.0 systems. The process is started, but "ntp -q" only shows the "LOCAL" > clock and none of the configured servers. In addition, manually running "rcntp > restart" needs to be done several times to get the daemon properly initialized. > Sorry, thats "ntpq -p" to show the daemon status...
Nothing major due it work here. Hopefully it is not a configuration problem. In case of a configuration this bug would be invalid and the support should be asked about.
please provide me with details on your ntp setup...
After further investigation, the problem appears to be caused by DHCP not assigning an address right away during the network setup. A message that DNS resolution is not yet available appears on the boot screen. An address does get assigned later on in the boot process, but NTP does not retry attaching to the configured time servers and drops them off the list requiring a restart of ntpd. Issuing "rcntp restart" does not restart the daemon properly the first time after logging in. If during the boot process the network setup does not "go into background" or a static IP address is assigned, no problem exists.
if you rely on the address to be there you have to use a static one of use other mechnisms then boot script dependencys. Stuff like pre/post scripts with ifup or something sorry..