Not sure if ping utility bypasses it, but uncheck "block RFC 1918" at bottom of WAN interface screen.

Also you need FW Rule to ensure traffic for 192.168.216.0 does not go via the WAN Gateway.

I just now got a LB2120 LTE modem plugged into OPT1with the same scenario. Gateway is online for main ISP WAN but offline for LTE WAN2. I can place in load balancing with both on Tier1 and WAN2 pulls data, I can connect it to the WAN by itself and it works just fine however it always is stated as the gateway offline. Have you by chance found a solution to your problem or an adjustment to the gateway monitor?

UPDATE:
First off, just figured out that LTE networks only use carrier grade NAT which means even if you place your LTE modem in bridged mode, you're still in their network and not truly going to pull a public IP. Ugh. Their network is an IPv6 with IPv4 network translation. I do get a IPv6 with some configuring but without majorly adjusting settings within pfSense and I need IPv4 anyway.

Now to answer your question, since we're still in a carrier grade NAT even in bridged mode, pfSense will say offline. Since the LTE modem is the failover and if neither WAN or WAN2 work then I'm toast I just set the WAN2 to not be monitored and failover works. Go to:

System -> Routing -> Gateways and edit WAN2 gateway and next to Gateway Monitoring check to disable.

Turns out it wasn't even a routing issue after all. I had a switch that was running in my patch panel that when connected to the network causes the packet loss, very odd. No idea why the new router fixed it temporarily. Guess I'll set pfsense back up.

@vernichter said in Pfsense on ionos with /32 subnet and gateway in other network:

a looped back NS message is dectected during DAD for fe80:xxxxxxxx. Anoter DAD probes are being sent

Check this thread :)
https://forum.netgate.com/topic/98857/a-looped-back-ns-message-is-detected-during-dad

depends on the dhcpd that is serving your leases..

LOL yes it was that my captive portal was blocking the entire thing.

Can be closed :)

You might be getting a multicast flood, When you connect the cable to switch C, do you have constant traffic on the indicator? If so, then your vlans are bridged at two points and a loop is created.

in the end i got it to work !

Match your traffic in firewall rules and switch the gateway (display advanced) to any gateway or gateway group you like.
0_1549392414639_pfSense_Policy-routing.png
https://docs.netgate.com/pfsense/en/latest/routing/directing-traffic-with-policy-routing.html

-Rico

@johnpoz thank you for your reply. I'm fairly new to pfsense and this was my first time using it in a real live scenario. I had suspicion about that being the reason, makes a lot of sense now, will be something I indeed will need to keep in mind :)

Port forwarding has ZERO to do with what IP pfsense has on its wan... If your public IP changes, then just use a dynamic dns setup so you know how to get to that IP.

Port forward on pfsense is just what traffic hits its wan IP, and were to send it on too.

thank you