How IBM plans to compete in the cloud

IBM hopes to reduce the complexity of moving workloads to the cloud through Cloud Paks,’ its series of specialized packaged offerings that enterprise can implement relatively quickly. But is it enough?

Data center in the clouds virtualization
Thinkstock

IBM has been pushing hard on being a competitive threat in enterprise cloud, but is far behind the leaders like Amazon AWS, Microsoft Azure and Google Cloud. It’s latest strategy to become more relevant, in addition to buying RedHat for its cloud expertise, is to develop a series of “easy on-ramp” Cloud Paks that it claims can significantly reduce the amount of time necessary for enterprises to be cloud-enabled. But is this enough to change the capability of IBM to compete in a highly competitive modern cloud environment?

While much hype has been produced about the rapid pace of enterprise cloud deployments, in reality we estimate less than 25 percent of enterprise workloads are currently being run in the cloud. That doesn’t negate the importance of the growth of cloud computing – but it does set some parameters around just how prevalent it currently is, and how difficult it is to move enterprise workloads to a cloud architecture.

Most organizations are at least experimenting with cloud workloads, but many also have a very mixed cloud environment. Of the organizations running cloud workloads, we estimate at least 80 percent have a multi-cloud environment that includes access to both on-prem and public cloud instances, as well as using multiple providers (e.g., AWS, Azure, Google, Oracle, IBM, SAP, etc.). This makes the world of cloud deployments very complex.

IBM has been pushing hard on being a competitive threat in enterprise cloud, but is far behind the leaders like Amazon AWS, Microsoft Azure and Google Cloud. It’s latest strategy to become more relevant, in addition to buying RedHat for its cloud expertise, is to develop a series of “easy on-ramp” Cloud Paks that it claims can significantly reduce the amount of time necessary for enterprises to be cloud-enabled. But is this enough to change the capability of IBM to compete in a highly competitive modern cloud environment?

While much hype has been produced about the rapid pace of enterprise cloud deployments, in reality we estimate less than 25 percent of enterprise workloads are currently being run in the cloud. That doesn’t negate the importance of the growth of cloud computing – but it does set some parameters around just how prevalent it currently is, and how difficult it is to move enterprise workloads to a cloud architecture.

Most organizations are at least experimenting with cloud workloads, but many also have a very mixed cloud environment. Of the organizations running cloud workloads, we estimate at least 80 percent have a multi-cloud environment that includes access to both on-prem and public cloud instances, as well as using multiple providers (e.g., AWS, Azure, Google, Oracle, IBM, SAP, etc.). This makes the world of cloud deployments very complex.

Betting on Cloud Paks

What IBM is focused on reducing the amount of heavy lifting necessary for companies to move their workloads to the cloud. This is being done through what it calls “Cloud Paks.” These are basically a series of specialized packaged solutions that can be implemented relatively quickly.

Many enterprises have simply shifted their workloads without optimization to the cloud environment from their traditional data centers, but that does not optimize the capabilities that cloud computing offers.

Rearchitecting for the cloud should include containerization of major application components in something like Docker, which can then be managed by an open sourced Kubernetes orchestration framework for optimization of resources and efficiency. We expect that containerization will ultimately be the defacto standard for running workloads in the cloud, and not just the wrapped up monolithic app implementations brought over from client server implementations.

IBM says that its recent acquisition of Red Hat gives it the primary mechanisms to accomplish its goals of containerization and orchestration of a wide range of cloud based workloads.

Indeed, Red Hat is the leading Linux-based provider of enterprise cloud infrastructure. It’s been adopted by 90 percent of enterprises and has more than 8M developers. Its OpenShift technology is a key component of its success, as it provides a way to easily deploy multi-cloud environments through a full stack control and management capability built on top of industry standard Kubernetes and deployed in a virtual Linux stack.

Leaders of the Paks

So far, IBM has created five Cloud Paks – Applications, Data, Integration, Automation and Multichannel Management. Each is built on a foundation of OpenShift and Red Hat Linux, but with “special sauce” that IBM claims makes implementing the paks simple for enterprise users.

These five special use cases will ultimately be expanded by IBM and will also be made available to the ecosystem for expansion by individual companies and/or vendors. And although these Cloud Paks are optimized to run on the IBM Cloud, because they are built on top of OpenShift they are able to run on virtually any cloud foundation, creating a no-lock-in solution that should be more palatable to companies who are not IBM-centric or exclusive.

The intent of Cloud Paks is to offer a pre-configured, containerized and tested solution that is certified by IBM. This approach is meant to eliminate a lot of the unknowns in deploying workloads in the cloud. While we think this is a great approach to simplification, there is still a significant amount of customization that needs to be made for each instance of the solution that will be unique to an individual organization’s needs. As such, a significant portion of the Cloud Pak deployment must be custom implemented by IBM services. That in and of itself is not necessarily a problem, but it does mean that this is not a simple “off the shelf” solution that can be implemented easily by internal IT staffs in most organizations.

Cloud deployment? Who ya gonna call?

Services add cost and time to deployment, but starting with a certified base level Pak can still lead to reduced time to market and less effort than starting from scratch. In that regard, we believe that IBM is on to something here. Many of its customers, particularly the largest enterprises in verticals like financial services, healthcare and government agencies, still look to IBM for its expertise as they move their workloads from their current data centers to a future cloud-implementation strategy.

Anything that can be done to allow companies to start with an “80 percent solution” like a Cloud Pak is advantageous. Of course, this applies only if the current workload fits into a Cloud Pak scenario, and with the limited number of Cloud Paks currently available, it may not.

Bottom Line: We applaud IBM for seeing the need to have more standardized approaches to deploying enterprise workloads in the cloud and to reducing the complexities of working in a hybrid, virtualized multi-cloud world.

Giving companies a starting point that contains all of the key technology components is far superior to starting from scratch. But until IBM has more Cloud Paks or creates a broader ecosystem to create Cloud Paks externally, it may have limited appeal. But built on open standards like Red Hat Linux and OpenShift, IBM has taken a credible step to becoming a key player in the future of enterprise cloud technology vendors.