Edge computing has gradually climbed the hype curve over the last couple of years, and it now stands at the center of why we do new things and launch new technologies. Why is it so important, what does it mean, where is the money behind the movement, and what does it mean to you? These are all good questions, and there is no simple answer to any of them.
Edge is what happens when we start to look at how we take advantage of all the computing capacity across networks and enterprises—the same way cloud has done in a data center—as a real problem to be solved.
What we are really talking about is being able to use this distributed and disparate infrastructure as a common resource for doing "the cool stuff," even though "distributed" and "disparate" are terms that a data center operator would try to avoid. Edge as a concept promises a lot, including heartache and difficulty along the way. It's not yet clear how far it can realistically go and how much benefit we can realistically expect from all this effort.
It's hard… so why bother?
What's the great opportunity? That's the big question. Why would I invest in edge—what is my ROI, who will be in control of this domain, is it limited to telecoms? Does it open opportunities for new players to enter what has traditionally been a closed and highly regulated market?
For now, these are open questions. We know, or believe, that if we had a ubiquitous infrastructure that was easily accessible and programmable, it would open new avenues for services, software, businesses, and communities to extend how we produce and consume services and understand their relevance to our immediate surroundings and environment. These might include:
- augmented reality in a venue or concert hall taking an extraordinary experience to the next level;
- contextual services as you walk or drive down a street;
- better, real-time in-home entertainment and security services; and
- improved reliability or simplification of brick-and-mortar retail services.
None of these likely compel the market to invest heavily in solving one of the most complex networking challenges of the digital age. But taken together, and with the promise of unknown opportunity and benefit, we have an almost unquestionable need to pursue the technology to what may soon be sci-fi levels of operability.
Moving beyond the hype
Most of the world now interacts with and connects to the internet via mobile phones and devices. Consumers have voted with their feet and moved their digital interfaces to mobile-based solutions, making this the first and most urgent target for a ubiquitous, open, accessible edge infrastructure.
Telecom network operators and service providers—the entire ecosystem—have quietly improved how consumers access and consume digital services. The next phase of this evolution is to establish this common global network as a platform for innovation, exploration, and new businesses.
By design, 5G is a mobile architecture with a focus on disaggregating infrastructure, enabling regional domains, and increasing bandwidth for our devices. This architectural slant establishes the opportunity for open infrastructures to take a fundamental role in how we interact with and consume digital media and services.Open infrastructure is absolutely necessary for this change to occur at the scale needed. It will be built by large communities of companies, individuals, and institutions striving to drive toward the next iteration of our digital age. This provides the scope for commercial entities to integrate with, adopt, and extend the ecosystem, further enabling new cycles to leverage what has come before.
Next steps
Today the spotlight is on the infrastructure layer. Where is the hardware and how do we expose that capability to an upper layer of applications and services? How do we leverage current data centers and deployments, and how do we migrate towards the envisioned future without breaking existing services? Solving this problem in a way that allows applications to extend their reach from a data center to a geographically diverse set of resources, where applications can leverage the location and context of those resources to address the hype, is where a lot of focus is today.
While there tend to be more questions than answers in the industry (as well as in the open source ecosystem around edge), we can't let this to stop us from working toward our goals in an iterative manner. The Edge Computing Group under the OpenStack Foundation umbrella is focusing on collecting use cases and identifying requirements across the industry, keeping the scope on the infrastructure-as-a-service (IaaS) layer. The working group's mission is to define reference architectures that fit different edge scenarios, acknowledging that there is no "one size fits all" solution. Instead, we need to focus on enhancing the building blocks to use them in different combinations to fulfill the needs of multiple segments of the industry.
While we are shifting between business models, it is crucial to iterate, starting from existing environments and evolving towards the realization of futuristic architectures. Telecom operators are continuously moving towards 5G. They're already running deployments to serve edge computing use cases where they are facing challenges on the infrastructure layer and the services' operation and lifecycle management.
Manufacturing companies are starting to utilize the fundamentals of edge computing to optimize the efficiency and improve the performance and reliability of their machines, as well as the productivity of the factories they ship to (for instance, in the textile industry). These companies also are transitioning to include software technology in their business models, in which they are relying on open source for building blocks as well as processes.
We are not moving away from the benefits of cloud computing, but taking them to the next level.
With this in mind, the focus is on enhancing basic functionality such as identity management, image management, identifying the needs for persistent storage, and exploring how we will combine technologies such as virtualization, containers, and specialized hardware on the edge.
Project teams in the OpenStack community are working on addressing current pain points, such as improving federation capabilities in Keystone, the OpenStack identity service and image caching in Glance, the OpenStack image service based on a reference architecture with a minimalistic view that builds on the immediate needs of the edge ecosystem.In parallel to design, development, and testing activities, we are iterating on the architecture to identify new requirements and gaps to address on the infrastructure layer while focusing on enhancements (versus reinventing the wheel).
As the demand for edge is extremely high, there are numerous groups in the open source and standardization ecosystem with different scopes—from the infrastructure to the applications to addressing different segments of the industry, such as telecom, automotive, or manufacturing. Collaboration between these groups is more crucial than ever.
The Edge Computing Group and the OpenStack community are actively working together with communities such as OPNFV, ONAP, and Kubernetes to ensure interoperability and smooth integration between the different components, as the success of edge strongly depends on combining the different technologies.
Ildiko Vancsa will moderate a panel discussion, Beyond the Hype: Edge Computing Working Group Update, featuring Beth Cohen, David Patterson, Ian Wells, and Christopher Price at the OpenStack Summit in Berlin, November 13-15.
1 Comment