RTP Servers
This screen is not implemented. It is hidden from the menu, and the form itself carries the notice “Not implemented yet”. This page describes what the screen is intended for and what it currently contains, so the entry is not mistaken for a working feature.
RTP is the protocol that carries the actual audio of a call, separately from the SIP signaling that sets the call up. Because audio is a continuous stream, it is far more sensitive to distance than signaling is: every extra network hop between the two parties adds delay and increases the chance of jitter and packet loss.
What a dedicated RTP server would be for
The purpose of placing an RTP server near a group of users is to keep their audio local instead of carrying it to a central site and back. That reduces one-way delay, reduces jitter, and takes sustained media load off inter-site links, which is the traffic most likely to saturate them.
This matters most when a single deployment serves several geographically separated offices. It matters very little for a single-site system, where media has nowhere else to go.
What the screen contains
If the screen is made visible, it holds a list of servers with three settings each:
- Server URL - the address of the RTP server.
- Cluster participation - whether the server is in service.
- Weight - its share of traffic relative to the other servers in the list.
Related documentation
- Servers - call-processing nodes in the cluster
- Cluster Configuration - how nodes are assigned roles
- Deployment Options - cluster architecture