Five years ago, the industry was debating whether the hybrid cloud was real or not. I was an early hybrid cloud supporter the past decade, but I rode the wrong platform horse, OpenStack. For mass deployment across different verticals and sophistication levels, OpenStack was not robust enough, and required a lot of work to stand it up and operate. The industry is now onto containers and bare metal.
Now, literally every major infrastructure and cloud vendor wants to be present across the edge to cloud, responding to the customer yearning to run workloads anywhere, from “Edge-to-Cloud” without meaningful changes. The only thing we’re debating is “multi-cloud”, or orchestrating different, branded, public clouds. Some vendors are in denial while announcing Kubernetes control planes.
The week before last, IBM announced the general availability of IBM Cloud Satellite to enable IBM’s hybrid cloud services anywhere – any cloud, on-premises, and at the edge. In this article, we delve into IBM Cloud Satellite and provide an edge use case.
Today’s reality: Hybrid cloud with multiple cloud providers
Hybrid cloud, the combination of private (either on-premises or hosted in a colocation facility) and public cloud infrastructure, is here to stay.
The hybrid cloud model, and the use of multiple cloud providers, are now the norm among enterprises. Adopting a multi-cloud strategy may be from a desire to avoid vendor lock-in or to take advantage of best-of-breed solutions or, in some cases, by accident as the result of ‘shadow IT.’
A multi-cloud strategy has many advantages, but it undoubtedly adds an extra layer of management complexity — mainly if multi-cloud adoption develops in an ad hoc manner rather than being planned from the ground up.
The promise of efficiency quickly becomes cumbersome to manage, with a lack of visibility of what workloads are running on the different clouds, creating problems when compliance and regulatory constraints are part of the business.
Multi-cloud success requires workloads that are portable across architectures.
Modernizing and building applications with Containers and Kubernetes orchestration is, to some extent solving the portability challenge.
Multi-cloud drives a need for an expansive and robust multi-cloud network infrastructure — extending from workloads to access — capable of delivering agility, flexibility, elastic scaling, operational efficiency, and security. Multi-cloud networking solutions must provide through software a consistent network policy across multiple cloud providers. Network configurations, security policies, troubleshooting, and even analytics and reporting should be available regardless of where the workload is running.
The last and most important critical success factor in a multi-cloud infrastructure is data reliability, data accessibility, and business continuity. Each cloud vendor has its architectural methods and ways of organizing data that are often incompatible with other clouds.
Without a comprehensive and seamless data management strategy for multi-cloud workload provisioning, management and de-provisioning, there is a risk of compromising or losing valuable data. Data must move fluidly and securely from anywhere to anywhere; be governed in a consistent, holistic manner; and be visible and controlled through a single platform.
IBM Cloud Satellite
IBM Cloud Satellite is an extension of the IBM Public Cloud that can run inside the customer’s data center or out at the edge. Over the past few years, IBM has ported all the cloud services on Kubernetes, allowing everything to operate consistently. IBM Cloud Satellite is no exception, running on Red Hat OpenShift, the Kubernetes management environment.
Each Cloud Satellite location is an instance of IBM Public Cloud running on Red Hat Enterprise Linux (RHEL) operating on any x86 hardware or an IBM Cloud Pak System, a self-contained turnkey private cloud prepackaged hardware/software offering. In the future, IBM will support Linux on IBM Power. I discussed in an earlier article Red Hat OpenShift on IBM Power Virtual Server.
Each Cloud Satellite location is connected using IBM Cloud Satellite Link which provides the connection to the IBM Cloud control plane. It provides audit, packet capture, and visibility to the security team, while a configuration utility provides a global view of applications and services. IBM Cloud Satellite Link provides a simple way to manage the connection between IBM Cloud and the satellite location with visibility into all the traffic going back and forth with control over what endpoints on both sides of the link.
A compelling catalog of capabilities
IBM Cloud Satellite extends the IBM Public Cloud providing access to IBM Public Cloud Services such as Red Hat OpenShift, IBM Cloud Databases, Continuous Delivery pipelines, AI, and others. It also joins a portfolio that includes OpenShift, both as a software framework available from Red Hat and as a managed public cloud service run by Red Hat on AWS or Google Cloud. For IBM Cloud Satellite, the first two services to be supported will be Cloud Pak for Data and OpenShift as a Service. IBM Cloud Pak for Data as a Service extends Watson Anywhere with a flexible, secure way to run AI and analytics workloads as services across any environment.
Development and IT teams can deliver software faster by creating a speedy, scalable software delivery pipeline with cloud-native tooling, deploying them to any location, scale them on demand, and improving the speed of development through service API-based consumption.
IBM collaborates with more than 65 ecosystem partners, including Cisco, Dell Technologies, and Intel, to build secure cloud services, helping clients run workloads in any environment via IBM Cloud Satellite. Infrastructure partners offer a choice of storage, networking, and server solutions to help clients leverage existing infrastructures to deploy IBM Cloud Satellite locations at data centers or the edge. Service partners plan to offer migration and deployment services to help customers manage solutions as-a-service anywhere. IBM Cloud Satellite customers can also access Red Hat OpenShift-certified software offerings on Red Hat Marketplace, deployed to run on Red Hat OpenShift via IBM Cloud Satellite.
Managing video analytics at the edgeLumen Technologies and IBM have developed a solution to use real-time video analytics to ensure worker safety. Video is a great edge use case. Using video analytics at the edge to decipher if someone’s wearing a hard hat or not. The entire software stack resides at the office building edge, reducing the latency so workers can be alerted immediately going into a dangerous area not wearing proper safety equipment.
The solution uses video cameras to send images in real-time to a video management server. IBM Video Analytics software quickly processes each image, triggering an alert (if needed). If the system operates more slowly, a person at risk could already be many steps into the restricted area before being stopped.
Customers using the Lumen platform and IBM Cloud Satellite can deploy data-intensive applications like video analytics across highly distributed environments such as offices and retail spaces and take advantage of infrastructure designed for single-digit millisecond latency. Because the application can be hosted on Red Hat OpenShift via IBM Cloud Satellite from the proximity of a Lumen edge location, cameras and sensors can function in near real-time to help improve quality and safety.
Use cases can be quickly adapted utilizing cloud-native practices, for example, in a COVID use case where you warn someone not wearing a mask or are not being worn correctly and checking temperatures with thermal devices.
IBM’s new solution brings a secured, unifying layer of cloud services to all environments, regardless of where data resides. Customers can now deploy cloud services securely in a multi-cloud environment for both public and private cloud, on-premises, or at the edge. For instance, IBM Cloud Satellite will support satellite locations on Amazon Web Services, Azure, and Google Cloud.
For me, this new offering from IBM checks all the boxes I discussed above.
- IBM has ported all the cloud services on Kubernetes, allowing for everything to operate in a consistent way
- A key differentiator will be the third-party ecosystem as it will allow Cloud Satellite to support highly specialized solutions currently absent from other vendors.
- IBM Cloud Direct Link provides a dedicated connection to the IBM Cloud that provides the administrative control plane. It provides audit, packet capture, and visibility to the security team, while a configuration utility provides a global view of applications and services.
- IBM is looking to offer data as one of the first services. Data services are elemental to many organizations or use cases; data needs to stay local for reasons ranging from latency issues to data residency requirements.
It’s great to see the progress IBM is making on the cloud which may be surprising if you had only tracked its rough start. It’s different after the Red hat acquisition. IBM has bet everything on the cloud. If the company isn’t successful in the cloud, the company will not have long-term success. Competition is fierce and this won’t be a cakewalk, but I believe IBM has some advantages in the most secure and regulated environments. Another big plus is that IBM has its own public cloud. Cloud Satellite is a nice addition to the IBM cloud story and albeit announced after many of its rivals, isn’t late as the hybrid market is quite new.
Note: Moor Insights & Strategy writers and editors may have contributed to this article.