Load balance not working properly

[hoanghien]

Hi,
I got 2 uplinks from ISP.
1. PPPOE with tag VLAN.
2. DHCP without VLAN.

I set up PPPOE as default gateway (lower priority).
Group gateway with 2 gateways as same tier.
Set up monitor gateway and DNS servers' IP per gateway.
Apply policy rule in LAN interface with group gateway.

Now when PC in LAN access website or ping, like half of connections are dropped.

I tried to use fail over group gateway, everything working good , ISP links are ok.
Do I miss any steps in configuration loadbalance?

[mscd]

Same problem here (OPNsense 21.1.7, two WAN-links, both with static/internal IPs as exposed hosts behind ADSL-routers, one link untagged WAN-port the other TAGGED VLAN, both in tier1, load-balancing) ... the interesting thing is ... disabling "sticky connections" (stickiness = 0) seems to solve the problematic behavior, but otherwise, I think this setting should be activated in the load-balancing case (https-session-problems).

At the moment I can't figure out how to get WAN-balancing properly working with "sticky connections". Any advice?

Best regards,
mscd

[Matzke]

I also have problems with gateway group.

Could you check if your problems are solved when you disable "shared forwarding"?

But just to say - when disabling it, your Trafficshaper etc. won't work properly. When you don't use traffic shaping or captive portal you could try to disable shared forwarding.

I see a similar problem in my configuration - gateway group and shared forwarding = on --> sporadic internet dropouts. When using default gateway in firewall rule or disabling shared forwarding it works.

https://forum.opnsense.org/index.php?topic=23460.15

[mscd]

Thanks a lot for your reply ... i did some testing with "shared forwarding" disabled (WAN-load-balancing with two WAN-interfaces, both in Tier1) and activated "sticky connections" feature ... and it seems to work now.

=> So I think there must be a bug in the implementation of "shared forwarding" in conjunction to multiWAN-balancing?!