ntp-client fails to run at boot (dns issue)

[danielittlewood]

I am not sure how to actually find the logs that I get shown at boot time, but the message is roughly "could not run ntp-client - temporary failure in name resolution". I can see NetworkManager starting just before it, and the script /etc/ntp-client has dns in its dependencies:

depend() {
before cron portmap
after net
use dns logger
}

although I don't know in detail how the init.d scripts work, so maybe this is not the right configuration.

If I run ntp-client after startup, it works fine and sets the time up correctly. Is there anything I can do to fix this?
_________________
With man it is impossible, but with gcc all things are possible.

[alamahant]

Hi
What is the output of

[danielittlewood]

[alamahant]

[danielittlewood]

That doesn't seem to have worked - I changed my ntp-client script to have

[danielittlewood]

Ok, here are the logs:

[danielittlewood]

That fixed it, but made my boot time substantially longer. Fixed, I guess?

[alamahant]

Maybe disable openvpn and only let NM handle the vpn connection.
Are you trying to connect to a vpn server?
_________________

[pietinger]

[figueroa]

EDITED and ADDED: I blew it. The logs shown below are from my desktop machine that syncs from my server without name resolution. I think on my server I've got lousy ntp servers in /etc/conf.d/ntp-client. I assumed that the gentoo defaults would work. Now I've got to experiment. I'll leave the entire post in case portions of it may be helpful. Otherwise, ,with regrets. If I discover anything, I'll post about it.

Are you sure it fails to start? I see the same error on-screen during startup, but it starts anyway. I can see that it started in both rc.log and the boot log.

Here is a segement from /var/log/rc.log-20220925.gz