Branch offices and stores, the manufacturing floor, remote sensors, ships and planes — all of these locations represent the vast and distant frontier known as “the edge.” The edge is not new. Those branches and locations have always been out there, functioning and supporting the business. But what is new are the state-of-the-art technologies that can help you take greater advantage of the edge than ever before.
What is edge computing?
The concept of “edge computing” means moving the compute resources from the centralized data center to or near the physical location of a user, a device, or a source of data. By placing computing services closer to these locations, users benefit from faster, more reliable services resulting in better user experiences. Meanwhile, companies also benefit by being better able to process data and minimize latency.
The conventional method for addressing the need for computing services at the edge is to employ proprietary solutions designed by specialty vendors. With the use of open source technology, however, you can now deploy compute resources and application services at the edge in a more cost-effective and practical way.
Edge computing use cases
The following are three primary situations where you would need application services at the edge.
The “operations edge” refers to cases where your company has operations or devices functioning at the edge, and you would like to collect and analyze data from those sources to support more intelligent decision making.
Traditionally in this situation, data would be collected at the edge and then sent to a centralized data center for processing. In many cases, however, transmitting terabytes of data results in unacceptable levels of latency. By collecting and analyzing the data locally, at the edge, you save compute resources and bandwidth, because you are not sending massive amounts of data back and forth to a centralized data center. Processing that data on location lets you process quantities of data that would otherwise be impossible.
Data can be analyzed to help organizations make better decisions, either through automated processes or manually. The information derived from data at the edge can help you improve a wide range of functions that impact the business, such as customer or user experience, productivity, preventative maintenance and fraud prevention, just to name a few.
Often, the advantage of operations edge involves analyzing data in near real time in order to help the organization make time-sensitive decisions. In these situations, it is even more important to perform the data collection and processing on site to minimize latency. In this way, the operations edge can lead to new levels of operational and business efficiency.
An example use case for operations edge is the shop floor in a remote manufacturing facility. By analyzing data produced by sensors built into the equipment, users can monitor performance and gauge whether preventative maintenance is required in order to keep the equipment from breaking down.
Achieving this requires an underlying platform that can unify the disparate systems generating data. Traditionally, manufacturing systems have been isolated from each other, so integrating all the systems under one platform is a challenging initiative. This particular use case would be well-served by a broad open source platform that is designed for this sort of integration.
“Enterprise edge” is a strategy to extend the application services in your enterprise to users and data in remote physical locations. These application services include functionality such as integration, messaging and stream processing, which will help modernize and support the applications you currently use on the edge.
Example use cases of enterprise edge are branch locations for restaurants, retail stores, pharmacies or banks. Each of these locations must be able to provide digital resources to support the business. In recent years, digital transformation has increased the amount of computing required at many branch locations, as an increasing number of previously manual tasks become digital. Self-service catalogs and checkout are just two of many examples.
By bringing application services to the remote branches, you can improve the customer and employee experience, enabling faster transactions and easier sharing of information.
Similar to operations edge, enterprise edge benefits from a uniform open source platform that is designed for integration.
Development at the edge
“Development at the edge” involves the ability to develop, deploy and update applications running in remote locations. This type of development can be challenging, especially when you need this capability across hundreds or even thousands of remote sites.
Applications deployed at the edge must be designed specifically for these situations. They must be lightweight and portable, while still providing all of the required functionality.
As an example, traditional Java is one of the most popular programming languages in the world, but not all the frameworks are ideal for applications on the edge.
Applications built on traditional enterprise Java architectures deliver high throughput, are resource hungry and function best when they have access to large amounts of memory and processing power. You need to develop much more lightweight applications for the edge, however, similar to the types of applications you build for functions on the cloud.
A framework like Quarkus can help you develop these sorts of lightweight applications for the edge. Quarkus lets developers use their existing Java knowledge and experience, providing the same Java frameworks they have used in the past.
Java developers are able to use Quarkus to build apps that have a faster startup time and take up less memory than traditional enterprise Java-based microservices frameworks. This translates to lower costs, because it takes less memory and processing power to run the same application.
A focus on open source and flexibility will help future proof your edge sites and help your developers create innovative applications that fully utilize and build on the edge advantage.
Edge computing challenges
The physical remoteness of edge locations poses several challenges that must be solved in order to keep the business running.
Unreliable connectivity: Since edge locations are so far away from the central data center, connectivity is often not strong or reliable.
Vendor lock-in: Traditionally the need for compute power and application services at the edge has been solved by niche vendors with proprietary solutions. While this may solve the initial problem, it creates a whole new problem of vendor lock-in. Your organization becomes dependent on that specific vendor, and you become locked into a product which may not provide all the features, capabilities or cost-effectiveness that you need. A flexible open source platform can help break the cycle of vendor lock-in.
Lack of technical expertise: Edge sites often have limited or no on-site technical expertise. It helps to have an infrastructure based on a flexible open source platform in place, so remote failures can be fixed by local non-technical personnel managed centrally by a small number of experts located elsewhere.
Nonconformity: Since edge sites are remote, the technology supporting the site often evolves independently, being implemented in different ways at each location. If a company has multiple edge locations, this can make management a complex, difficult, costly and time-consuming task. A uniform edge computing platform is needed to develop universal site management processes across all your edge computing sites.
Overhead cost: Scaling out servers to multiple sites on the edge is typically more complicated than adding the same capacity to a single central data center. The increased overhead it takes to support remote sites can be complicated and costly.
Physical security: Physical security of edge sites is often much lower than that of core sites. An edge strategy must account for a greater risk of malicious and accidental security threats.
How to choose an edge solution
We've established that it's best for an organization to deploy a universal edge computing platform in order to make edge locations more functional. When choosing an edge computing platform, we suggest that you look for the following capabilities:
Universal platform: Most importantly, your organization needs a common, horizontal, unified platform — from the core to the edge — providing a consistent development and operations experience.
Open source: For the most cost-efficient approach, your edge computing platform should be based on proven open source technologies, rather than proprietary niche solutions. This allows you to avoid vendor lock-in and enable the platform to be integrated with new cutting-edge technologies.
Interoperability: Your edge computing platform should be interoperable with a wide variety of components sourced from various vendors, again to avoid vendor lock-in.
Consistency: You should run a consistent deployment model from the core to the edge. You can achieve consistency best by using the same tools and processes you are already using in your enterprise infrastructure. This includes automated provisioning, management and orchestration of hundreds — if not tens of thousands — of sites that have minimal or no IT staff.
Flexibility: Your edge computing platform should offer flexible architectural options to best meet your connectivity and management needs.
Scalability: The edge can consist of tens of thousands of devices and sites that can be constantly changing in a dynamic environment. The ability to economically scale up and down is a critical requirement for your edge computing platform.
Availability: On the edge, one of your top priorities is to ensure that sites continue to operate in the face of network failures and unreliable connectivity. Your edge computing platform must make continuous operations possible at all your remote edge sites.
Automation: Your edge computing platform must be able to automate processes across your edge sites, similar to the automation built into your enterprise.
Security: The ideal edge computing platform should provide a strong security posture to help protect data and infrastructure in vulnerable edge environments.
Red Hat Application Services for the edge
Red Hat delivers a complete edge computing solution to help you meet all the challenges of the edge, delivering the must-have capabilities listed above while also helping you avoid vendor lock-in.
Red Hat OpenShift is designed for the hybrid cloud, and we see edge computing as an opportunity to extend the hybrid cloud all the way to the data sources and end users. In addition, Red Hat’s portfolio of enterprise software is optimized for lightweight deployment in any location — whether on-premises, in the public cloud or at the edge.
Red Hat Application Foundations is a connected set of application services, combined with OpenShift, designed to help accelerate containerized application development and delivery across hybrid cloud, multicloud and edge environments.
This toolkit provides developers with the following advantages:
Self-service tools and services
In addition, Red Hat Application Foundations offers ready-to-implement components including:
High-performance data streaming services
Application integration connectors and transformation components
Red Hat is uniquely positioned to break the hold proprietary vendors have on the edge, addressing the needs of organizations extending their infrastructure and services to the edge through Red Hat platforms and application services. Our edge computing solution can also be managed using the same tools as your enterprise infrastructure.
Red Hat's advantages at the edge
The bottom line is that Red Hat's edge computing portfolio extends the following advantages to your operations at the edge.
High performance: Delivers low-latency, highly available applications to the edge where connectivity is not always available or consistent.
Application stability: Provides stable, reliable and consistent services at the edge.
Greater control: Enables you to take greater control over your edge sites that have minimal or no staff, allowing you to remotely manage your IT assets at widely distributed locations as if they were physically part of the enterprise.
Near real-time data analysis and decision making: Brings you closer to your edge data than ever before, making real-time data collection, processing and analysis possible at your farthest edge sites. Ultimately, this intelligence lets you make decisions in real time to address a wide range of issues that directly impact your business.
Lower cost: Reduces your operational costs at the edge by breaking the chains of vendor lock-in.
Improved user experience: Perhaps the most important advantage of Red Hat's edge computing platform is its ability to improve your end-user and customer experience, based on all the previously mentioned advantages.