mistakenly pressed sent earlier, here is complete query.
If we want to use multiple kamailio servers for WebRTC based user agents, what would be the best practice:
1) WS/WSS Outbound/Edge Proxy
In this case would it not become a single point of failure? e.g. if we have a wss edge proxy, would registrars under it be on plain SIP? if so this WSS Edge Proxy should handle media then ? or the nodes should handle ?
2) Each UA mapped to a specific SIP URI and then routing of calls based on shared DB socket id (to the registrar server of that UA) - it is not a generic approach as each proxy will have its own address, making client side to be restricted to use one proxy only.
Looking forward to some directions here. Thanks