Bringing the Cloud to the Edge

Bringing the cloud to the edge

Cloud computing has revolutionized the way we access and process information throughout our personal and professional lives. “The cloud” –on-demand services delivered by a shared pool IT resources and accessed by users via a web browser— offers computing (e.g. Amazon Web Services’ EC2, Azure’s VM, Google’s Cloud Engine N1), storage (e.g. AWS S3, Azure Files, Google Cloud Storage), networking (e.g. AWS’s VPC, Azure VNET, Google Cloud CDN), and many more applications. This is powered by an infrastructure of many interconnected data centers located in strategically planned regions (the “Regional Cloud”).  

However, data center locations are by nature very limited, while the users and devices who need to access these critical services are anywhere and everywhere. As user experience and latency requirements become more stringent, this poses a new challenge to the established infrastructure. To get as close to the users and devices, the solution is “Edge Cloud.”

Edge and Regional Cloud, Working Together

Edge Cloud’s ability to provide cloud-like capabilities at the infrastructure edge, i.e., close to the user or devices, has made it an emerging trend in network architecture today. It is a seamless extension of a Regional Cloud and can be hosted in edge data centers deployed at the infrastructure edge. In that way, its intention is not to replace Regional Cloud entirely; instead, it is a complementary solution for just a few specific, latency-sensitive applications. The following chart maps some of those applications, especially those further enabled by 5G service.

Unlocking commercial opportunities from 4G evolution to 5G

Source: GSM Association, “Unlocking Commercial Opportunities, From 4G Evolution to 5G”

In addition to applications, another distinction between Edge Cloud and Regional Cloud is the computing tasks they perform. While Edge Cloud is used for data pre-processing/processing, data caching, request responding, or service release, more advanced computing tasks— the deep learning of AI/ML, for example— will still be performed in the Regional Cloud.

New Model, New Challenges

Edge Cloud implementation is not without its challenges, however. It requires an enormous amount of edge data centers, likely tens (if not hundreds) of thousands of locations. This proliferation of data center locations could become a major network management challenge. Furthermore, Edge Cloud by its nature has significantly more security risk vectors than a well-protected Regional Cloud. As the number of edge data centers increases, the number of attack points will inevitably increase. In order to mitigate these new threats, edge data center operators will have to consider the specific defense mechanisms necessary to protect these critical applications and computing tasks.

To learn more about how open network solutions can support Edge Cloud implementation, contact us today.

Share on facebook
Share on twitter
Share on linkedin