The two prominent performance-limiting factors associated with an Session Manager server are Session Manager server processing occupancy and memory constraints.

Session Manager processor occupancy is theoretically directly proportional to the rate at which Session Manager initiates and tears down SIP sessions. A SIP session consists of the signaling associated with a connection between two SIP trunks, communicating via a Session Manager instance. Session Manager SIP sessions required per call for various call types indicates that there is not generally a one-to-one correspondence between call and SIP session.

As a design criterion, the total occupancy (including the static occupancy) of the Session Manager server complex should not exceed 80%. That number is analogous to the 65% static + call-processing occupancy used in designing Communication Manager systems. The reason the design threshold is higher for Session Manager than Communication Manager is that Session Manager requires no processing cycles to be reserved for hardware maintenance activity.

Memory constraints establish a second, independent constraint, pertaining to the maximum number of simultaneous SIP sessions and the maximum number of TLS sockets supported by a single Session Manager instance.