Join this talk if you want to get a good understanding on what the OpenSIPS clustering support is and how it works. Starting with the architecture and requirements of the communications system as a whole, the presentation will continue with the runtime behavior for the cluster. This will teach you what to expect from an OpenSIPS cluster, how to dynamically resize it and how it behaves in the eventuality of cross-site network link failures.

In the second half, we will dissect a natural use case for the clustering support: a distributed SIP user location service. There are a couple of interesting problems associated with this type of service (NAT traversal, SIP contact pinging, etc.), both of which become a lot more straight-forward to solve using an underlying clustering engine.