Edge computing is a buzzword that has been floating around in the cloud computing industry. Over the years, this computing paradigm has been steadily evolving and gaining momentum as an innovative solution to streamline the flow of traffic from the Internet of Things (IoT) devices and to deliver real-time local data analysis.
Defined by IDC as “a mesh network of microdata centers that process or store critical data locally and push all received data to a central data center or cloud storage repository,” edge computing allows for efficient and real-time data processing without latency. As a result, businesses can increase network performance and reduce the amount of data at risk while allowing businesses to expand their computing capacity through a combination of IoT devices and edge data centers.
Given the efficiency, scalability, and security benefits that the edge provides, knowledge-centric businesses are embracing this distributed computing paradigm to gain a competitive advantage, and leading the way is Mirantis, a California-based company distributor of OpenStack.
SourceForge recently caught up with Nick Chase, the Head of Technical and Content Marketing at Mirantis, to talk about public and hybrid clouds and edge computing and discuss the launch of their latest product offering that is purpose-built for edge cloud use cases. Eckert also shares how the Kurbenetes-based Mirantis Cloud Platform Edge (MCP Edge) enables virtual machines and containers to run at the edge of the network.
Q: Catch us up since we last spoke. What has Mirantis been up to?
Nick Chase, the Head of Technical and Content Marketing at Mirantis
A: Since last year, Mirantis has continued focusing on enabling Network Functions Virtualization (NFV) and cloud-native technologies. Part of doing that has been adding Kubernetes as a core open cloud platform at the center of Mirantis Cloud Platform alongside Mirantis OpenStack. We’ve also launched the beta of Mirantis Application Platform, based on Spinnaker, which empowers developers to deploy containerized applications while maintaining control of compliance and security.
Q: Cloud computing has come a long way in ten years. A decade ago, it was largely the preserve of technology-based sectors. But today, the cloud is well and truly open for business. In your opinion, where does the future lie, in the private or public cloud?
A: In a word, yes. The reality is that in order to really take advantage of the digital transformation that the cloud provides, it is increasingly necessary to build hybrid platforms and multi-cloud applications. The advantages of the public cloud, such as flexibility, potential cost savings, and operational simplicity still hold, but beyond a certain scale it does not make sense economically — and that is if the practical and regulatory aspects of the application do not require it to be on-premises. Private cloud is not going anywhere anytime soon; the reality is that many workloads simply can not — or should not — be run in a public cloud environment.
The ideal solution is an architecture that enables you to confidently and reliably build your architecture across multiple cloud environments.
Q: What are some of the challenges that some of the early adopters of public cloud have faced?
A: Many of the challenges early adopters of public cloud have faced are the same as those private cloud practitioners have faced, only more so. Cloud is a different mindset, and it is one thing to create applications and services that are insecure within your own datacenter; it is another thing entirely to do it on hardware you share with anyone and everyone.
In addition, the public cloud introduces additional challenges around latency on the technical end of the spectrum and compliance on the non-technical end.
Q: How does Mirantis support hybrid and multi-cloud strategies?
A: Mirantis Cloud Platform has been designed from the beginning as a multi-cloud solution, enabling the reliable creation of infrastructure that spans multiple architectures and providers. Its “Infrastructure as Code” foundation and monitoring and auditing capabilities make it possible to easily replicate infrastructure in different environments, and to know what is deployed where, and by whom.
Q: What exactly is edge computing and how does it work? What makes it unique from cloud computing?
A: Edge computing is more about requirements and use cases than overall architecture. For example, both cloud and edge are decentralized architectures, but where cloud computing assumes that the pieces close to the user are sending data back to the central cloud for processing, edge applications typically can’t tolerate the network latency this involves, putting more of the work on the “edge” node itself, as it’s closer to the user. Also, because edge hardware is closer to the user (think edge nodes on a cell tower) the hardware has to be both more robust and remotely manageable.
This type of infrastructure enables you to build applications that provide capabilities such as real-time interaction and immersive experiences (including self-driving vehicles), locally specific content and functionality, and operator functions such analytics (including stripping the data actually sent to the core to the essentials so as to not overwhelm the system) and Network Functions Virtualization (NFV).
So you can think of edge computing as cloud computing with some additional requirements.
Q: We’ve learned about the capabilities of Mirantis Cloud Platform (MCP) in our previous interview. So what’s new with MPC Edge?
A: Mirantis Cloud Platform (MCP) Edge is a “tuned stack” for MCP. In other words, it has the same basic architecture as a generic MCP cloud — DriveTrain for lifecycle, Stacklight for analytics, and so on — certain decisions have already been made in order to start you off with a cloud that is appropriate for edge applications. As with all MCP clouds, you can then customize this model for your own needs, but you’re starting out with a huge head start.
Q: What does MCP Edge do? Can you share some use cases?
A: The MCP Edge architecture enables you to satisfy use cases such as immersive experiences (4K video 360 video streaming), real-time interaction ( e.g. AR/VR and connected cars), self-contained environments (e.g. stadiums and concerts that provide Internet access), compliance (e.g. geo-locking content or functionality) and analytics (e.g. performing initial aggregation on the edge node before sending data back to the core cloud).
For example, a surveillance system could be set up so that object detection was done in the edge node, and only “relevant” frames of video were sent back to the core cloud, significantly decreasing both network traffic and load on the cloud.
Q: What rising trends, strategies, and/or technologies do you think will affect the future of edge computing? How is Mirantis meeting these head-on?
A: In many ways, the edge is the “future of computing”. More and more functionalities are getting pushed out to devices rather than the cloud, and many of those devices are going to be introduced. For example, Amazon recently launched a new Alexa-enabled microwave just to prove it could be done and to encourage developers to bring its voice assistant to more devices.
All of this requires incredibly flexible infrastructures, which is why Mirantis developed DriveTrain, which looks at infrastructure as code, enabling flexible, secure, and intelligent management of that infrastructure.
Q: What’s on the horizon for Mirantis? Are there are any new products/developments your customers can expect before the year ends?
A: Following on the heels of the recently announced MCP Edge, this month Mirantis is set to launch the latest version of MCP, which provides operators a complete software solution designed for edge cloud use cases.
Highlights of the latest MCP release includes Kubernetes 1.11, OpenStack Queens, and OpenContrail 4.0. It also includes updates to the DriveTrain component upgrade pipeline, Kubernetes Upgrade Pipeline, granular OpenStack Ocata to Pike Upgrade pipeline, as well as documentation experience and security improvements in all areas.
About Mirantis
Mirantis is the leading commercial distributor of OpenStack. Established in 1999, Mirantis has been helping more than 200 enterprises and service providers develop and operate some of the largest open clouds in the world. Their current customers include iconic brands such as AT&T, Comcast, Adobe, State Farm, STC, Reliance Jio, Volkswagen, Vodafone, and Wells Fargo. The company specializes in delivering cutting-edge open cloud infrastructure to global enterprises using Kubernetes, OpenStack, and related open source technologies.