dhcpcd-discuss

út 10. 11. 2020 v 13:46 odesílatel Roy Marples <roy@xxxxxxxxxxxx> napsal:
>
> On 10/11/2020 12:17, Jan Pohanka wrote:
> >> So you've already said where you should look.
> >
> > Sorry for my misunderstanding, now I got it. You are absolutely right.
> > I have had eth0 interface specified in /network/interfaces with dhcp
> > option, I just haven't realized that this is the thing. ifup in
> > busybox is a binary not a script so it was not obvious at first for me
> > that it is starting dhcpcd.
> >
> > Our application used a command ifdown eth0 && ifup eth0 to reload
> > configuration or restart the interface when needed. What would be a
> > correct alternative using dhcpcd? dhcpcd -n?
>
> dhcpcd -n will cause the interfaces to load a new configuration and rebind.
>
> Here's the important bit - if you want to limit this to one or more interfaces
> specify them on the command line.
> If only one interface and dhcpcd is not already running in master mode then
> dhcpcd will start for that one interface only and could be the cause of the
> issue we are talking about.
> You need to ensure that dhcpcd is started in master mode only first, or not at all.
>
>
> > Can you also tell me what would be a correct approach to test if a
> > dhcp request failed and fallback static was used? I need to recheck
> > this periodically.
>
> In /etc/dhcpcd.exit-hook you could write this:
>
> echo "profile = $profile" >/tmp/"dhcpcd-$ifname.profile"
>
> And then look in that file to see the last profile used.
>
> Or you could just take some action in the exit hook script based on the profile
> used.
>

I think I got it all.
Thank you for your support and have a nice day.

Jan