Cisco Cisco Unified Contact Center Enterprise 9.0(2) Leaflet

This chapter presents an overview of the Unified CCE network architecture, deployment characteristics 
of the network, and provisioning requirements of the Unified CCE network. Essential network 
architecture concepts are introduced, including network segments, keep-alive (heartbeat) traffic, flow 
categorization, IP-based prioritization and segmentation, and bandwidth and latency requirements. 
Provisioning guidelines are presented for network traffic flows between remote components over the 
WAN, including recommendations on how to apply proper Quality of Service (QoS) to WAN traffic 
flows. For a more detailed description of the Unified CCE architecture and various component 
internetworking, see 

Cisco Unified CCE has traditionally been deployed using private, point-to-point leased-line network 
connections for both its private (Central Controller or Peripheral Gateway, side-to-side) as well as public 
(Peripheral Gateway to Central Controller) WAN network structure. Optimal network performance 
characteristics (and route diversity for the fault tolerant failover mechanisms) are provided to the 
Unified CCE application only through dedicated private facilities, redundant IP routers, and appropriate 
priority queuing.

Enterprises deploying networks that share multiple traffic classes, of course, prefer to maintain their 
existing infrastructure rather than revert to an incremental, dedicated network. Convergent networks 
offer both cost and operational efficiency, and such support is a key aspect of Cisco Powered Networks.

Beginning with Cisco Unified CCE Release 7.0 (provided that the required latency and bandwidth 
requirements inherent in the real-time nature of this product are satisfied), Cisco supports Unified CCE 
deployments in a convergent QoS-aware public network as well as in a convergent QoS-aware private 
network environment. This chapter presents QoS marking, queuing, and shaping recommendations for 
both the Unified CCE public and private network traffic.

Historically, two QoS models have been used: Integrated Services (IntServ) and Differentiated Services 
(DiffServ). The IntServ model relies on the Resource Reservation Protocol (RSVP) to signal and reserve 
the desired QoS for each flow in the network. Scalability becomes an issue with IntServ because state 
information of thousands of reservations has to be maintained at every router along the path. DiffServ, 
in contrast, categorizes traffic into different classes, and specific forwarding treatments are then applied 
to the traffic class at each network node. As a coarse-grained, scalable, and end-to-end QoS solution, 
DiffServ is more widely used and accepted. Unified CCE applications are not aware of RSVP, and the 
QoS considerations in this chapter are based on DiffServ.

Adequate bandwidth provisioning and implementation of QoS are critical components in the success of 
Unified CCE deployments. Bandwidth guidelines and examples are provided in this chapter to help with 
provisioning the required bandwidth.