Use Case and Reference Architecture

The reference architecture is based on a modern Carrier Ethernet MAN that takes into consideration the transformation that is required to facilitate diverse new services, applications, and use cases. Some common principles exist to deliver Layer 2 and/or Layer 3-enabled services for point-to-point, point-to-multipoint, and multipoint-to-multipoint solutions with more intelligent mechanisms that enable the coexistence of L2 and L3 services and improved high availability models. The architecture is referred to as Cloud Metro but carries several important characteristics in the amalgamation of service and content providers. These shifting industry trends demand massive bandwidth and increased service scale while also supporting more complex metro workloads.

A major goal of Cloud Metro is the adaptation of cloud principles into metro networks. This comes in the form of systems that support a sophisticated feature set, including the array of EVPN technologies, SR-MPLS/SRv6, and the ability to support inter-domain traffic engineering or seamless architectures across disparate networks. It must include capabilities to support and integrate the services and solutions that are found in traditional metro networks. This is a differentiating factor that characterizes the requirements for supporting x-to-anything connectivity models or building infrastructures that become access agnostic while also blending with virtualized network functions and devices.

Metro networks can vary between service providers, but the design principles are largely consistent. In the traditional metro network, the design focuses on supporting north-to-south traffic patterns where services are backhauled across access, aggregation, and core network segments and centrally aggregated. The costly scale-up architecture is supported by resilient modular systems with dense feature sets that can carry a significant failure blast radius. New challenges emerge with the growth of edge cloud complexes, leading to massive subscriber traffic increases and exasperated by the consumption of expensive links and ports while degrading the customer experience. A new design is required.

As illustrated in Figure 1, with the emergence of a new model moving to the right, aggregation nodes evolve into a lean edge role, with certain tactical and strategic advantages realized as the traffic patterns are better contained within the metro ecosystem. East-to-west traffic flows are cost-optimized and a significant reduction in the failure blast radius. Additional benefits are realized with reduced power consumption and lowered cost-per-gigabit while improving scalability and the customer experience.

The ACX7000 family is ideally positioned to support the lean edge role with an advanced feature set capable of serving a majority of customer requirements and providing critical interconnectivity points for these new cloud complexes. In parallel, the MX Series multi-services edge component serves the crucial role of managing more complex interconnectivity, service stitching attributes, Pseudo-Wire Headend Termination (PWHT), or high-scale BNG use cases.