IBM’s decision to acquire Red Hat for US$34 billion is a bold stroke, an admission of weakness and a mighty complication.
It’s a bold stroke because Linux has become the most common OS in public clouds. Even Microsoft Azure is now more than half Linux and Red Hat dominates those installs.
IBM’s emphasis of hybrid clouds as the reason for the deal shows the company thinks it will be better off tending to workloads that users put in clouds other than its own.
Which is where the deal shows IBM’s weakness because its own IaaS remains unsophisticated compared to rivals and efforts to renovate it have struggled. By effectively buying itself a cloud that spans several IaaS rivals, IBM shows its own IaaS ambitions have dimmed.
Which may not be a bad thing: IaaS is just someone else’s computer and doing it well requires huge infrastructure investments and operational rigour. Google, MIcrosoft and Amazon must make those investments and develop those skills to support their core businesses.
IBM, however, is not a natural fit for IaaS because so much of its value proposition is tied up in its proprietary platforms and software. The latter can be monetised by piping them into other clouds as native services,which is what IBM and Red Hat have announced they plan to do.
Buying Red Hat to give it management tools across multiple IaaS clouds, and probably an easier route to market for its software delivered as SaaS, therefore means a chance to turn IaaS weakness towards IBM’s strengths.
And the complication? The deal is being pitched as making IBM and Red Hat the premier hybrid cloud company. Which is just what VMware says it already is. Yet IBM and VMware are partners, with vSphere-powered clouds being one of IBM’s bigger cloud successes.
Yes, plenty of the VMs running under vSphere on the IBM cloud will use Red Hat, and lots of IT titans offer overlapping and/or competing products. But there's a difference between a little overlap and buying a direct rival that IBM will have to explain to VMware.