Sponsored By: Pluribus Networks
Resources that have traditionally been deployed in centralized data centers are beginning to be deployed in new locations closer to where traffic is being generated in order to improve latency and service quality and reduce transport costs. This move toward the distributed cloud, as it is called, is often part of a larger digital transformation that enterprises and CSPs are undertaking to improve operational efficiency and competitive positioning. Distributed cloud infrastructure is necessary because many new applications and services are impractical to deliver from a centralized cloud. Latency, which is particularly important in applications like AR/VR and connected cars, is the clearest attribute necessitating distributed infrastructure. Minimizing the amount of data that must be transported back to a centralized data center will also be important in managing costs, given the expected explosion of data these new services will generate. Some services will require autonomous operations in remote locations. Meanwhile, privacy and security concerns – many of which are driven by new regulations – will force more data to be handled and stored locally.
The advent of 5G is helping to drive demand for distributed edge architectures because of the technology advancements it promises. Network performance, including speed and latencies, will improve dramatically, as will densities and handoff speeds. Combined, these will make new services practical for the first time. Network slicing, whereby separate environments can be created to serve individual customers or deliver specific services, will become more prevalent, giving CSPs the opportunity to more easily support multi-tenancy. The servers, storage, and networking devices used to build the distributed cloud are also evolving. Disaggregated solutions help CSPs both reduce their CAPEX through lower hardware costs and more efficiently deploy features now that software is no longer tied to hardware. Software-defined networking (SDN) brings easier network automation and service orchestration to improve operational efficiency. The applications riding on those platforms are moving to new cloud-native architectures – microservices-based and delivered in containers – making them more suitable for distributed cloud infrastructure.
Download your copy now of this Pluribus Networks White Paper, exploring distributed cloud and how centralized data center resources are now being deployed to new locations.