I have a customer who has a public ip assigned to his router from us that is connected 4 hops away on a all bridge network design. When the issue occurs, which could be at any time they see their pings jump from a standard of 30-50 ms to a rapid increase to 300 - 400 ms that will last for about 3 pings then settle back to the normal. This is causing their voip phone calls to have a lot of jetter and there is really no pattern.
Data is not an issue at all and not affected in any way, just the voip calls, so I would like to isolate them in their own vlan back to the edgemax router, until such time, I can convert the network from an all bridge network to a routed network. Or setup QOS on the router to handle the voip traffic for this customer.
In the near future I would like to get away from the all bridge network, as this is not the way to operate the network now that it has scaled to 200+ units across many different towers / locations, but that will take some design considerations and carefully planning.
I'm hoping that putting them on their own vlan or setting up QOS for the voip traffic will correct the issue, any help, comments, and other suggestions would be a great help.
All equipment is Ubiquiti, and all switches are tough switches. The route taken to the customer is as follows,
Edgemax router v1.7 -> connected to Cisco switch - > AP on Building -> Tower1 -> TSwitch - > Tower2 -> TSwitch - TSwitch -> Tower3 - TSwitch -> Customers Cisco Router
Building AP to Tower 1
Tower 1 - SU -> TS -> AP
Tower 2 - SU -> TS - TS -> AP
Tower 3 - SU -> TS - AP -> Customers SU Radio -> Cisco Router
Radios are Nanobeam 5AC 19 on Building to Tower 1 running beta 8.0 8.3.0538 both sides
Nanobeam M5 400 from tower 1 to tower 2 running 5.6.2 both sides
Nanobeam 5AC 19 from tower 2 to tower 3 running 7.2.3 both sides
Rocket 5AC PTMP on tower 3 to customer which is a Nanobeam 5AC 19 running 7.2.3 both sides
All Toughswitches are running latest software for these units.
Thanks