PFSense doesn't switch back to primary connection in multi-WAN setup
Sorry, but I find many entries with that problem, but no real solution.
I have 2 WANS, usually all traffic via Wan1. If Wan1 has high latency or loss, switch to Wan2. This works so far so good. Wan1 is Tier1, Wan 2 is Tier 5.
BUT: It does not switch back to WAN1, if it is ok again...
Is this a bug that has not been solved since years??? I paid nearly 800 euros for my applicance, so it would be nice if that would work without any issues...
snore last edited by
If you paid, perhaps contact support directly via email? I'm sure they are there for you.
they are not sure if it is a bug or a feature...
Still waiting for a solution. My message from today:
Hi Azamat, the router (WAN_DREI_LTE) is turned off for purpose at around 02:00 MEST for about one half hour, but then reboots. The netgate should switch to the redundant WAN (WAN_POST_AT) for that time, and then back to WAN_DREI_LTE. It does not. Today the Gateway was even shown as down (red) in the dashboard, although it was up (because I connected via VPN through this WAN). Can you please add at least a major bug, as this behaviour is a big problem. My redundant WAN for example is a slow connection and only for cases the fast WAN is not working properly (fallback). Regards, Gunther
I hope that they find a solution for that, it can be quite severe if you are using a metered or slow connection as fallback...
@hebein hi! Just one question. Did you specify different DNS server record for each WAN interface?
i have nothing set there, should I?
i have an own dhcp server service in the lan for clients.
@hebein yes, you can find it here https://docs.netgate.com/pfsense/en/latest/routing/multi-wan.html#dns-considerations
You can use your ISP's DNS record.
ok, thanks for the hint. tomorrow I will know if it was the solution ;)
If yes, I owe you.
Renat last edited by Renat
@hebein you're welcome! I hope this will solve your problem. Let me know pls)
snore last edited by
Did it work?
Sorry for the late reply. It worked 4 of 5 times by now. So seems to be the solution. :) Thanks @Renat
@hebein glad to help you! I think log analyzing will help to reach 100%)