Is there a route in the kernel to <upstream resolver> that works when
the default route is not present?
the @ppp2 syntax sets SO_BINDTODEVICE on the socket used to talk
upstream, but that's not going to work unless there's a route.
Simon.
On 30.12.2025 07:04, Luigi Baldoni via Dnsmasq-discuss wrote:
Hello,
before stating the obvious, this happens when server=<upstream resolver>@ppp2
is set.
On my system I have concurrent ppp connections, each used independently.
If ppp0 is down or a default route through it is not set, dnsmasq won't forward
queries to
the upstream resolver, even though it seems to use only ppp2 in normal
conditions. At least
that's what tcpdump shows.
Any idea what's happening? I could try using uid-based policy routing I
suppose, but I don't
get why this would be necessary.
Regards
_______________________________________________
Dnsmasq-discuss mailing list
[email protected]
https://lists.thekelleys.org.uk/cgi-bin/mailman/listinfo/dnsmasq-discuss
_______________________________________________
Dnsmasq-discuss mailing list
[email protected]
https://lists.thekelleys.org.uk/cgi-bin/mailman/listinfo/dnsmasq-discuss
