Two days ago I set up dual wan on my AX 92u, utilizing the failover feature with auto switchback. I’m using T-Mobile HI gateway as the primary isp input and Fios as the standby. seems to be working correctly after two tests, both where I would unplug the primary isp wan cable to see how long it took to recognize the failure and make the switch To the backup. I didn’t time it, but it seemed to take 15-20 seconds. Plugging the primary wan cable back in triggered the router to switch back to the primary. I guess a couple of questions I have are…
Is this a reasonable way to test this?
Are there other routers that make the switch from primary to backup more quickly? For example, are there routers that would recognize the outage and switch almost instantly?
Fore those interested in using Dual WAN rather than failover mode, where you want failover in the event of a WAN problem too.
I have the Asus T-AX86U with latest Merlin FW, and two WANs which I use in load balance mode. IPv4 only.
My Primary WAN is BT internet (FTTC) via modem (no double NAT). I consider it primary because it has lower latency. My secondary WAN is Virgin cable via Docsis modem (no double NAT). Higher latency but much higher bandwidths. I have a gaming PC main interface routing to the primary WAN, everything else routed to secondary WAN, including a second interface on the gaming PC which is used to upload (Twitch) streams.
It's not really load balancing, as all my outbound traffic choice of WAN is determined by the routing rules, which are:
192.168.50.163/32, all, Primary WAN
192.168.0.0/16, all, Secondary WAN
(I used a /16 as a mega catch all. No harm done.)
Firstly, unplugging the Primary WAN from the ASUS causes the secondary WAN to go down too (inexplicably...) for around 40 seconds. After that all devices including the gaming PC use the secondary WAN no problem. Reconnect Primary and it comes back into service immediately, and the gaming PC traffic is routed over it immediately too.
Unplug the secondary WAN cable next. Primary WAN stays up. All traffic is routed over primary after about 30 seconds. Reconnect secondary cable. All secondary traffic reverts back to it after about 30 seconds.
This basic failover and recovery is pretty OK for my needs, but I appreciate that it could behave very differently (worse) in the event that a different network issue occurs upstream from my cable-pulling (which others here have rightly pointed out).
tl:dr: Asus router load balanced WAN failover and recovery behaviour with routing rules in place does appear to work ok for cable pulls in a basic test.