Few things are worse than your backup solution not working when the primary fails. If you use Direct Routing to connect Teams to the PSTN, you likely have two paths for these calls. If you don’t, it’s something you should consider!

If you own your own SBCs, these may be located in two different offices, or hosted in two different Azure regions, or some combination of the two. If you use Direct Routing as a Service where your carrier provides the SBCs, they like have an “east coast” and “west coast” type of configuration.

OutBound Calling

Within Teams, you are in control of your outbound call flow. You can set your Voice Routing Policy to include one PSTN Usage with both SBCs listed for a “round-robin” scenario:

Or you may have your Voice Routing Policy contain two PSTN Usages, each with a single SBC. The PSTN Usages are ordered, giving you a Primary-Backup relationship (you could add more, but let’s Keep It Simple, System administrator!):

If you have the first scenario, your outbound calls will round-robin between the two paths and you’ll know the backup path is working. If you have the second scenario, you would need to periodically test the backup path by taking down the primary SBC. Yuck, nobody wants to do that, plus if you’re using DRaaS the provider isn’t going to do that for you (nor can you trigger it by deleting the SBC config from Teams, but double yuck, why would you want to do that!)

What you can do is create a second Voice Routing Policy that has the PSTN Usages listed in the opposite order. If you have users in cities that broadly match the “east/west” or whatever geography of the SBCs, you can assign this policy to the other group of users. If not, you could select a few IT staff.

There’s a catch with both of these options. While they’ll send traffic over both of your paths when everything is working, they’ll also happily send traffic over just one path should the other go down, and NOT tell you. You’d need to add other services or monitoring packages to help here, so what’s an admin to do?

The Solution

The solution here is to setup two additional Voice Routing Policies, each with one usage and one SBC. These get assigned to your IT staff, or to some other staff members who regularly makes but can tolerate things breaking. These staff are your “canaries in the coal mine”, and they can alert you when a call path is unavailable.

Inbound Call Paths

For inbound call handling, you are at the mercy of your PSTN provider. They will need to configure your inbound calls for round-robin, or assign numbers to have alternate SBC priorities.

Is this solution perfect? No, not by any means. The biggest flaw here is that there is no alerting (apart from the canary users). While monitoring and alert is ideal, the ideas outlined here can ensure that all of your call paths are regularly used and are ordinarily fully functional.