Service Provider Infrastructure Today
In the last one decade, service providers have deployed a mix of physical and virtual infrastructure and built data centres on centralized cloud computing architectures. Technology enablers have matured and problems of scale, elasticity and speed are addressed.
Evolving Applications, Services and Requirements
With the world order evolving over the last one decade, machines are progressively transforming from pure work horses to the intelligent ones. Industry 4.0, V2x, Autonomous car, smart city infrastructure, AR / VR, Video surveillance and gaming applications are some of the technologies that have exposed us to enormous amounts of information and event data in the field. Just as human neural network has got local intelligence in the spinal reflex arc, local processing power and local decisioning is necessary to reduce response time, latency, cost of bandwidth and improve quality of experience.
To address the change in demands of these new applications and services, infrastructure is required to undergo transformation. It is time and cost efficient to process data locally where it is getting generated i.e., the edge of the network, which means centralized cloud architectures of yesteryear needs to be transformed.
Edge Infrastructure is the solution to transport and process the voluminous data at scale. Private 5G connectivity to industry 4.0 and campus, vehicle-to-everything (V2x) use cases demand low latency and quick response times.
Challenges Faced by Service Providers
5G is set to shake up the service provider landscape and bring in new order to the service provider infrastructure. With all these changes, edge computing doesn’t remain a choice, it becomes a necessity.
To support industry 4.0 devices for their application needs, 5G infrastructure is required to offer speeds that are at least many folds higher than 4G.
Real-time processing demands on Edge need new network and compute infrastructure. With higher degree of distribution of compute nodes,operations costs will go up. Business Leaders are to invest on infrastructure management and orchestration solutions that are light weight, able to manage edge infrastructure at scale, and cost efficient.
As a large number of Edge nodes get deployed, footprint of the edge nodes needs to be revisited to optimise TCO of the nodes and ROI.
As compute gets pushed to the edge, Security risks increase since the net attack surface widens. Hence new generation of security solutions that secure the compute and network will need to evolve. New security requirements need to be addressed between Data centres and Edge nodes.
Operational practices of yesteryear data center deployments are out-dated and do not suit the distributed edge infrastructure. There is a need to evolve a hybrid approach to preserve investments on traditional operational solutions and adapt new generation operations solutions.
Several open source communities are rapidly evolving infrastructure components to support containerized workloads. The rate at which the enablers are evolving holds promise. Service providers will likely migrate from virtualization to containerisation in the next few years.
Multus kind of plugins enables PODs to have multiple network interfaces. DPDK and SR-IOV helps to support line rate at POD level. IPSec acceleration supported by QAT is promising. Kubernetes enable the clustering of containerized workloads across Edge infrastructure. Kubernetes also helps to containerize legacy virtual machines.Service mesh technologies such as NSM and Linkerd are evolving to support service function chaining use cases. Kubernetes community is also evolving lightweight orchestration components such as Kubeedge to support nodes with low footprint. Kubeflow is evolving to support AI pipelines at the Edge. Theses enablers hold promise to shift workloads on top of Kubernetes clusters.
Service providers can offer B2B2X services and increase revenues by reducing latency between service end points and increasing QOE.
With processing power and applications shifted to Edge, enterprises will save on bandwidth costs.
With 5G edge infrastructure, a number of new services such as video surveillance and VR can be offered at affordable tariffs.
Going deep into the Edge value chain, Service providers can offer IaaS and PaaS services to enterprises and consumers that demand high bandwidth and rapid service response. This is a space where service providers have a clear edge over public cloud providers.
Service providers will be able to offer Augmented Reality (AR) and mobile gaming applications with improved responsiveness of applications. It holds promise to increase revenue.
Content delivery networks will leverage the capabilities offered at the edge and save on transport costs.
While Edge infrastructure holds promise to increase ROI to service providers and QOE to enterprise customers and users, it also brings down IT operations costs.
About ACL Digital
ACL Digitalis a forerunner in offering product engineering services to both communications service providers and equipment manufacturers. With more than a decade of experience in cloud hosted NFVI and cloud native infrastructure space, ACL Digitalis helping communication service providers transition from yesteryear Virtualized NFVI to Kubernetes based NFVI.