Redundant WAN

The first option is still to leverage redundant WAN connections between a branch site and a datacenter. In this model a media gateway and Mediation server typically are deployed in the branch, and users are homed to a Front End pool that exists in another site, as shown in Figure 32.4. This enables the Mediation Server to locally communicate with the gateway and use RTAudio when interfacing to other Lync servers across the WAN connection.

Figure 32.4. Offices with redundant WAN connectivity.

Although it is technically possible to not locate a Mediation server in the branch, the local gateway would have to send a G.711 audio stream across the WAN, which is more likely to result in audio quality issues.

This is also a popular model for scenarios with centralized SIP trunking services that are delivered to a datacenter because there is then no local gateway or PSTN connectivity in the branch. In that scenario the only way to ensure that those users can make and receive calls during a WAN outage is to use redundant WAN connectivity.

The disadvantage to this approach comes in the cost of monthly fees for the extra WAN connection, which is not generally used unless there is an outage. Organizations are paying an operational expense to provide redundancy during an outage of the primary WAN circuit. Also, if both WAN circuits are disconnected due to physical issues then users will have no services at all.