I have a two Pepwave system setup. MAX BR1 MK2 and a BR1 MINI connected to the BR1 MK2 via the WAN port. I do this to have two cellular modems active at all times and not have the expensive/lengthy switchover between two carriers and also allowing the use of both carriers per outbound policies. I’ve tried various policies such as Weighted Balance or using the default in the unit which I believe is the lowest latency?
AT&T is in the primary (BR1 MK2) and Verizon is in the BR1 MINI, we’ll call it the secondary unit. The problem that I see is the primary unit always favors and choose Verizon as the network when both AT&T and Verizon are in Priority 1. I believe the problem is that the outbound policy testing is seeing the faster hope between the two units, i.e. near zero latency between primary and secondary Pepwave’s instead of measuring the latency to an Internet resource. AT&T may be significantly faster and Verizon unusable but Verizon is always the one selected when both share Priority 1.
Not sure if this is a bug but hopefully my use case can indicate a potential change needed in how latency is measured to determine the best WAN source.