Route10 WAN Failover - Not restoring correctly

I have just bought a Route10 - overall very sleek device!

I am having a strange problem that I am not sure how to resolve in relation to WAN failover.

WAN1 - PPPoE with VLAN tag
WAN2 - DHCP

I have setup metrics on both WAN connections, 100 for WAN1, 200 for WAN2.

When disconnecting WAN1 - the failover succeeds and all devices switch to WAN2. However, when reinstating WAN1, some of my devices stick to WAN2’s connection, but others do restore to WAN1. This issue persists after disconnecting/reconnecting the LAN cables, even swapping between ports has no resolution - the only resolution to get all devices back to WAN1 is to reboot the Route10. The issue is repeatable.

I am running firmware v1.4f

Sounds similar to this: Still Having Issues with Failover - #6 by diceman725

1 Like

I gave up for awhile. Continue to monitor this board to see if problem is addressed. Really like and Route10 and would like to come back, but this is a deal breaker for me.

1 Like

Hmm, it almost sounds like the connection tracking table needs to be cleared so the device connections drop for a second and start connecting to the proper WAN interface again.

As a test, maybe you could try the same test you described but log into the terminal by shift+clicking on the router and run conntrack -F to clear the connection table? Kinda like so:
Recording at 2025-06-21 16.22.35 (minus the initial misclick lol)

It looks like that’ll drop your connection via the Alta controller as well, but closing your tab and reconnecting should get you back. Then you can see if everything starts using the correct WAN again. If it does then it might be useful until there’s an official fix.

1 Like

Thanks, I gave that a go, but it didn’t help unfortunately, still stuck on backup WAN

Dang, I had hoped that would do something. Otherwise as far I as I can tell there’s an option to add to the configuration file that will automatically do that as well, but it might not be worth doing if the connection tracking flush didn’t seem to do anything.

That all being said, if you’re curious here’s the bit I pulled from the OpenWRT mwan3 documentation:

I think the configuration file can be modified by going back to the terminal and typing nano /etc/config/mwan3 then that option could be added under the wan interface. Although I haven’t had a chance to test it out myself!

cheers! Given the conntrack didn’t help i don’t think that will fix it, im on the fence of returning it tbh - on top of faulty WAN failover, I have since discovered there is no encrypted DNS support, nor RFC 4638 for PPPoE connections, no ability to change MTU either. I understand Alta is new so some things are expected, but I don’t see these things being addressed anytime soon

We’ll see whether this is corrected on a future release. That post is well over five months old to put into perspective timelines. It’s certainly a huge disappointment, especially coming from another manufacturer that had these features already in place and operated as intended.

1 Like

I have found a fix!

/etc/config/mwan3 - change sticky to 0, then restart mwan3 interface

Looking in to a way to stop it from reverting at reboot though

CLI command:
uci set mwan3.default_pbr.sticky=0
/etc/init.d/mwan3 restart

2 Likes

Might be able to add it to post-cfg.sh script so it’s loaded on startup, although I haven’t done that myself. Some examples of what others have used it for are listed here: Post your sanitized post-cfg.sh

Preferably there would be an option in the UI at some point, but it’s cool you found a fix!

1 Like

Tried that, the route10 wipes everything upon reboot :frowning: