BOSTON – IBM is adding a Cloud Foundry layer for developers that want to run container platforms on top of Kubernetes and the IBM Cloud. The move is targeted at helping enterprises gain more control over their container deployments and resolve some of the challenges of working with Kubernetes.
The Cloud Foundry platform is an open source, multi-cloud application platform-as-a-service (PaaS). It’s governed by the Cloud Foundry Foundation.
Described as an “experimental offering,” the integration has the Cloud Foundry platform running on top of Kubernetes, which itself runs on IBM’s cloud architecture. The integration provides a level of abstraction for developers that work on containers but don’t want to deal with the intricacies inherent in Kubernetes.
During an interview at this week’s Cloud Foundry Summit in Boston, Don Boulia, general manager for IBM’s Cloud Platform, said the deal will solve issues enterprises have when using IBM’s Kubernetes and cloud infrastructure. Companies need control over where they put their workloads due to geopolitical or regulatory concerns; and they need a level of isolation to guarantee that their content is not mixing with content from other companies.
“In our public cloud we did not offer a way to really mimic what an enterprise could get from their own on-premises or data center,” Boulia said. “It was more of a shared tenant model and only in some regions. This lets us address those issues.”
The Cloud Foundry integration also allows an enterprise to focus on building applications for production rather than dealing with Kubernetes. Boulia said Cloud Foundry is a good option for cloud-native enterprises as it offers a level of abstraction to ease management.
“We basically take care of all the Kubernetes stuff running underneath and the Cloud Foundry people can just run on top of that and get those applications up and running,” Boulia said. “Cloud Foundry leverages the container concept, but provides for a developer experience on top of that. It’s not just screwing around with Kubernetes, but about building applications.”
The move also continues IBM’s push to integrate different distributors on top of its Kubernetes and cloud infrastructure. Boulia noted similar efforts with SAP and SUSE.
“These are all different spins on what we are doing with these distributors,” Boulia explained. “The common goal is for better integrations. We think it’s better to get them running side-by-side if not completely integrated than having them run on totally separate tracks.”
Kubernetes and Cloud Foundry
IBM last year rolled out its initial support for the Kubernetes platform.
Boulia acknowledged that while Kubernetes has achieved production levels of stability, it can still be difficult for some developers.
“Kubernetes gives a lot of control and detail into the inner-workings of a container, while Cloud Foundry masks a lot of those details, but focuses on the production environment,” Boulia said.
The addition of Cloud Foundry on top of Kubernetes also helps IBM deal with questions from users that might be confused as to the need for the two platforms.
“What we don’t want is for an enterprise to come to me and say ‘which one should I use: Kubernetes or Cloud Foundtry?’” Boulia said. “These are really apples and oranges. They are tools for different purposes.”
The move also seems to resolve a support gap between IBM and Cloud Foundry. James Governor, analyst and co-founder of Redmonk, explained in a blog post last June that IBM’s Bluemix platform was “increasingly oriented to Kubernetes rather than Cloud Foundry.”
“IBM will support Cloud Foundry for existing customers, but will be shifting future investments to Kubernetes, and technologies such as OpenWhisk and Istio,” Governor noted. “IBM’s Cloud platform technologists are very, very happy with Kubernetes performance, particularly in terms of resilience, so far.”
IBM earlier this year launched a cloud analytics platform running on Kubernetes across its cloud infrastructure. That followed a move to include native Kubernetes support through a managed service model running on bare metal cloud infrastructure.