IBM today made generally available a managed Cloud Satellite edge computing service, based on Red Hat OpenShift, that now includes more than 65 partners, including Cisco, Dell Technologies, Intel and Lumen Technologies.
At the same time, IBM reveals that it is making IBM Watson Anywhere, an implementation of its AI software on Kubernetes, available as part of its IBM Cloud Pak for Data as a Service portfolio. That portfolio of services spans a range of databases and middleware based on Kubernetes that can now be deployed on an edge computing platform in addition to a public cloud.
Jason McGee, IBM vice president and CTO for IBM Cloud, says the ultimate goal is to make it simpler for organizations to deploy application code on highly distributed edge computing platforms. In many cases, it’s not feasible for internal IT teams to physically provision resources on their own, McGee says.
IBM Cloud Satellite is based on a Red Hat OpenShift platform for building and deploying applications that is itself based on Kubernetes. However, end customers of the IBM Cloud Satellite service will not have to manage Kubernetes, says McGee. Instead, end customers can opt to have IBM manage the entire environment on their behalf, or co-manage applications while IBM maintains management responsibility for all the underlying IT infrastructure.
Underneath the Red Hat OpenShift platform is an instance of Red Hat Enterprise Linux. IBM Cloud Satellite enables teams to define a location populated by a group of RHEL hosts. There is also Satellite Mesh, a federated instance of the open source Istio service mesh that can span multiple Kubernetes clusters, and Satellite Link, a tool for automating the administration of application-level firewalls.
IBM also provides access to a Satellite Config tool to provide a global view of applications and control over configurations and the deployment of applications.
In general, more code is being deployed on edge computing platforms as organizations look to process and analyze data at a point that is as close as possible to the point where it was collected. That approach minimizes the amount of data that needs to be transferred over a wide area network to a legacy, batch-oriented application. There will always be a need to share data across platforms, but McGee says where and how data is processed and analyzed – at the edge, on public clouds and on various other types of platforms in between – is rapidly evolving.
IBM is clearly trying to make Red Hat OpenShift the foundation upon which those hybrid clouds are built, regardless of who manages them. In fact, McGee notes, the term cloud computing no longer refers to a specific place where workloads run, but rather an operational style of managing the workloads, wherever they run.
The pace at which IT teams will be making a shift in the months and years ahead will, of course, vary. The challenge now is determining to what degree IT teams want to manage both the applications and the underlying infrastructure that make up those hybrid cloud computing environments.