We’ve been hearing about edge computing as a trend for the better part of a decade now. Edge computing was supposed to be this emerging computing paradigm that processed data closer to where it was being generated, enabling a range of networks and devices to process data at greater speeds and higher volumes. Heath Thompson, president & GM at Quest Software, explores why the promises haven’t quite been met yet and how the edge could live up to expectations in the near future.
With edge computing, it was always about the possibility of making greater use of “big data” (a term we rarely hear now) for AI, new types of applications, and greater efficiencies. However, there has not been significant traction for edge computing except in specific, well-known use cases. As network speeds have increased, and cloud and SaaS infrastructure has become more robust, ubiquitous and secure, the advantages of edge computing have right-sized in their importance.
Edge computing works by moving the compute source closer to the data. It makes sense to do this with things like IoT, and we can all appreciate iPhones and wearable IoT devices like watches that can process data without an internet connection and in real-time. In general, edge computing’s benefits are about response time, cost savings, data aggregation and consolidation, privacy and reducing threats of security breaches. In many locales, data sovereignty is also a key concern; organizations cannot, by law or policy, allow data to move outside certain domains. While this is critical, public cloud providers have solved for this with the geographic distribution of their data centers. It remains important as a driver for some edge computing but is not universal.
Navigating Complexities and Cost of the Edge
While the concept of edge computing is sound and the benefits of processing data close to the source are undeniable, the reality is that this approach also brings with it a higher amount of complexity and cost of management. As with any other part of an organization’s infrastructure, edge computing platforms also require management, maintenance, and security. Organizations already struggle with the scale of their networks, IoT devices and users on the network, and edge computing adds to this burden. Edge computing becomes another “node” on the network, which will require patching and opening new attack surfaces for threats like ransomware and data breach.
Advances in Technology Will Continue to Hinder the Concept of Edge Computing
Simply put, edge computing isn’t simplifying anyone’s life—it’s potentially making it more difficult. Beyond just the security and manageability considerations, edge computing can also add complexity to data governance and compliance. The data processed by edge computing platforms might have local storage of personal information (PI) data, subject to privacy controls like GDPR or CCPA. Also, edge computing may perform important data transformations, which need to be managed and governed as part of data lineage considerations in an overall governance strategy.
None of these points are meant to be eliminators for edge computing, but rather to call out the obligations that organizations must take on when they decide to adopt edge computing. There’s a dance of opposing forces here: the ever-increasing volume of data from our ability and desire to measure everything, vs. the need to produce actionable results from the data, vs. the challenges and complexities of managing edge compute infrastructure, vs. the quality and availability of other compute approaches. We are indeed implementing edge computing in some critical use cases, but we also realize that edge computing is not as cheap and cheerful – and thus widely deployed – as we thought it might be.
See More: Last Year’s Predictions and 4 Kubernetes and Edge Trends to Watch
If Not Edge Computing, Then What?
Edge computing is a bit of an anachronistic term if truth be told. Modern businesses don’t really think of “compute” as the way to architect for growth in 2023. Rather, organizations are envisioning their future outcomes and putting DATA at the center of their consideration. Organizations are building out data meshes, or data fabrics, as the nervous systems that will drive business operations.
Data meshes make data more accessible and available to users, directly connecting data to those using it: data owners, data producers and data consumers. They also give organizations better decision-making power, enabling teams to generate data while creating usable data products for other teams. Data meshes resolve problems like data bottlenecks more globally across the enterprise; where edge computing does it locally, a mesh does it from anywhere that makes sense. It can connect cloud applications to sensitive data that resides with a customer safely and securely. It can create virtual data catalogs from sources that can’t be centralized and can give developers the ability to query data from a variety of storage devices without access problems.
The technology follows four key principles that offer a lot of the promised but never delivered-upon business benefits of edge computing: domain-oriented decentralized data ownership; data as a product; self-serve data infrastructure as a platform; and federated data governance. Most importantly, a data mesh puts data where it belongs: at the center of any strategy enabling the growth and transformation of modern business.
Shifting the focus from compute to data is essential and also right-sizes the consideration of edge computing or cloud computing. Organizations today are focused on data, and compute is a means to an end on how to achieve results from the data. We’ve long since acknowledged that we’ll live in a “hybrid” world, where it’s not either/or, but both/and. But modern business is now rightfully focused on its data, and empowering organizations to use their data is where we need our focus to be in 2023 and beyond.
How are you shifting focus to the edge and beyond for leveraging data? Share with us on Facebook, Twitter, and LinkedIn.
Image Source: Shutterstock
