There are three elements that, by working in concert, will be instrumental in creating a successful edge computing strategy: edge computing, edge networking, and edge data management. Individually these three play very distinct roles, so it is important to understand how they can complement each other to unlock the potential of edge.
Defining edge computing
Containers and serverless infrastructure have allowed computing at the edge to become a reality. Previously, attempts to move powerful computation closer to users and machines to improve application performance and optimise server resources were hampered by a lack of highly distributed infrastructure. It is now feasible to run workloads across globally distributed Kubernetes clusters or to use serverless functions that reside in service provider environments. Cloud and content delivery networks (CDN) can be used to dynamically support distributed applications.
A streaming provider, for example, may use a variety of cloud resources or take a multi-CDN approach to deliver consistent, reliable performance in specific regions. Similarly, if it plans to stream a major new release, it could spin up a workload via a co-location data centre in a specific geography to support temporary high demand. The challenges for IT teams using edge computing tend to arise around data synchronisation and orchestrating global internet traffic and workloads.
Defining edge networking
The last two years have seen increased investment in infrastructure that allows employees and end-users to access services to support remote work and connectivity. As people have become geographically more distributed, data and network traffic have had to be routed dynamically across broader footprints to connect these audiences with applications. The key to this is edge networking.
To ensure the application workloads are successfully distributed, IT teams can use application steering policies, which allow traffic to be balanced between resources as conditions change. Edge networks also exist as part of physical infrastructure, supporting satellite offices, for example, where equipment should be maximised without increasing infrastructure. This is achieved through network management and connecting application resources.
Defining edge data management
Optimising data mobility is essential if organisations have large databases and computing workloads that run in multiple locations. It is a challenge to shift parts of the database to selected locations and expect it to deliver maximum performance and low latency without incurring the cost of distributing it across an international footprint. Addressing this will further enable scalable innovation at the edge.
The building blocks for edge success
IT decision-makers must have a clear understanding of each of the edge building blocks as current trends, including the support of a distributed workforce, and the increased demand for connectivity and online services, force companies to consider improvements in their infrastructure.
As an example, some large tech companies, such as content provider Netflix and collaboration service provider Dropbox have successfully built their own edge networks by combining distributed edge compute footprints with edge networking and data management strategies. They have invested in solutions that allow them to orchestrate their application traffic while automating the lifecycles of supportive resources. This has meant that they can deliver and control capacity close to their audiences and guarantee a reliable user experience.
In the medical diagnostics field, the finely tuned latency advantages of an edge network are accelerating diagnoses and treatments, enabling healthcare IoT devices to process vast volumes of data and supporting network speeds so that doctors can connect remotely and effectively with patients – a service that during the pandemic has become not only useful but lifesaving.
IT leaders have moved from cautiously investigating the opportunities afforded by edge technologies to specifying, ordering, and maintaining highly distributed edge networks and applications with outstanding results. Slowly but surely, edge strategies are being formed by organisations using the three elements of edge computing, edge networking, and edge data management, which together deliver an edge footprint that enables dependable, secure, and highly performative applications.