Multi WAN related issue possibly bug

[opns-sc0]

Following the manual to configure two WAN links for Load Balancing I am consistently running into an issue:
https://docs.opnsense.org/manual/how-tos/multiwan.html

As soon as I complete "Step 4 - Policy based routing" and hit "Apply" I get flooded with errormessages on the console:

Code Select Expand

arpresolve: can't allocate llinfo for <WAN IPv4> on vtnet1

It does not happen when I select "default" or any one of my existing WAN (v4) Gateways in the Gateway section. But whenever I select  the WAN Gateway Group defined in Step 2 of the guide, the error messages start coming.

I've seen some similar error messages reported in older threads but no resolution.

Any help would be appreciated!

[Shoresy]

What IP's are you using for gateway monitoring? Check Systems > Gateways > Single - generally it's recommended that you use a public DNS IP for monitoring, such as 1.1.1.1. Each gateway needs a different IP assigned, so your 2nd WAN could have 1.0.0.1 for example.

In System > Settings > General what are you using as your DNS servers assigned to each Gateway? The DNS IP's should match up with what you used for each gateway for gateway monitoring.

I see the arpresolve: can't allocate llinfo for <WAN IP> as well, but from my experience they're usually benign and WAN access is working fine from both gateways. The arpresolve messages will generally pop up if there are brief instances of packet loss on either gateway, which does happen occasionally.

Also try rebooting your OPNsense device to see if that helps cut down and/or eliminate the llinfo messages. Sometimes my OPNsense appliance will go a few days without any of those arpresolve messages, other times they seem to occur more frequently.

[opns-sc0]

I use different IPs for each of the Gateways, yes. (1.1.1.1 and 8.8.8.8)

DNS used in System > Settings > General are not the same but an internal DNS server that does work properly.

It was not only a few messages but over a minute or so filling the whole screen, so I don't think it's a problem of individual packets getting lost though.

Reboot also didn't make a difference.

I had it also tried on two different setups. One using lot's of VLAN and the other one only with physical interfaces. In both cases I had the same symptoms. After the multi WAN config according to the guide, I was flooded with these error messages.

[Shoresy]

I've observed the same behavior with multi-WAN, which is what I currently have on my own setup. Occasionally I see the arpresolve messages on my console, but it's not frequent. All I know is that it seems to happen more often after an initial config, or following an update, then it settles down and becomes more intermittent/sparse after OPNsense has had a few days of uptime.

I was itching to resolve this as you are currently...but after researching the arpresolve "failed to allocate llinfo on <wan interface>" for days, I found overall that it probably wasn't worth any further concern. Seems that the arpresolve behavior has been around for several years (on pfSense as well) and it relates to anything from NIC driver issues to WAN flapping. In my case, it doesn't seem to coincide with connectivity or performance issues, it seems to be related to one WAN interface attempting to resolve something on the other WAN interface (something related to dpinger), and vice-versa. Also, what may resolve this issue for some users doesn't work for others.

[wide]

I'm having this exactly same issue as Shoresy described.

Immediately when I switch the gateway to WAN Gateway Group it console is flooded with these

Code Select Expand

arpresolve: can't allocate llinfo for <WAN IPv4> on igc1 / igc2 messages.

For me it looks that <WAN IPv4> it mentions in interface igc1 llinfo message is the default WAN gateway for igc2 and similarly <WAN IPv4> it mentions in interface igc2 llinfo is the default WAN gateway for igc1.

In my case both WAN interfaces igc1 and igc2 get the IP configurations from two different ISPs DHCP servers.