Hello,
My use case: I have a mobile vehicle with 3 different LTE ISPs with different sized data plans. I would like to prefer the highest data capacity plans UNTIL the cutoff latency is exceeded and then move the next priority group.
I’m trying to leverage cutoff latency on a speedfusion vpn to suspend the link when it violates the threshold and have it move to the next priority group. I expected to either find cutoff latency under the wan interface properties or else have it apply outside the same priority group.
My configuration is as follows: I have 4 connections on a balance 20x connecting back to a fusion hub. I’m going to manipulate the facts here a bit to simplify for explanation purposes.
On the balance 20x - My interfaces are all in the same priority group.
WIthin speedfusion Cloud config managed and created by InControl2
I have
LTE #1 ATT - 100 Gig Data Plan(Priority #1 configured) - Cutoff 200ms
LTE #2 Verizon - 50 Gig Data Plan (Priority #2configured) - Cutoff 200ms
LTE # 3 TMobile - 10 Gig Data Plan (Priority #3 configured) - Cutoff 200ms
In a perfect world - I’m driving down the highway, and I’m using the LTE #1 plan. Life is good but congestion or a weak signal occurs and I would like it to suspend the ATT connection as it is at 500ms and have it fail over to the Verizon connection at 50ms until ATT becomes healthy/fast again. My problem is unless link 1 fails, it won’t move on to link 2, and if I put them in the same priority group, I’m just leaving it up to the algorithm to “pick the best” link(s) for DWB. It doesn’t understand or factor in the data plan capacity so quickly exhausts #3 just becuase it was the lowest latency/highest quality. (It’s worth mentioning I already use the feature to suspend my links and notify when capacity limits are being approached or exceeded, but that doesn’t solve my use case). Is there a way to have latency factored in to health checks or else have cutoff latency apply outside a SINGLE priority group within IC2?
Seems like a common scenario to just it prefer a given link until a latency threshold is exceeded, then have it fail back once it recovers.
Thanks very much! I’ve been digging through all the speed fusion docs and forum posts and can’t find anyone trying to do this across different priority groups. Everything just expects and applies to all the links being in the same priority group (in which case it works just fine - just doesn’t meet my use case of factoring in plan capacity). On the whole, speedfusion is amazing. Thanks!