By buying Red Hat, IBM just bought itself a clue in hybrid cloud computing. And it just might work
For those that haven’t been paying attention, 2018 is the year of open source. This may seem to be a bit of a strange claim given that open source has been with us for decades (and never did get its “year of the desktop”), but in a world where software matters more than ever, graybeard companies like IBM and Microsoft are paying billions of dollars to reinvent their futures through developers.
The biggest deal of all, of course, and one of the biggest tech deals ever, is IBM’s $34 billion acquisition of Red Hat. IBM CEO Ginni Rometty is almost certainly wrong to suggest the deal “resets the cloud landscape” and will “change everything about the cloud market.” It doesn’t. But what it does is give enterprises a massive reason to pause on full public cloud plans and double down on the hybrid cloud.
IBM gets to try again in the cloud
IBM has talked up its transformation for years, but those same years have seen dwindling revenues, dwindling profits, and dwindling relevance. In every market that IBM says matters most (AI, cloud, etc.), it’s an earnest also-ran. From Watson to SoftLayer, IBM has overpromised and underdelivered.
Red Hat, meanwhile, went all-in on Kubernetes, turning the container orchestration engine into its ticket to cloud relevance. While Red Hat lacks a public cloud offering to compete with an Amazon Web Services or Microsoft Azure, it has set up a fast-growing OpenShift business that gives enterprises a way to navigate their multicloud, hybrid environments.
Red Hat was doing exactly what IBM had hoped to do. Instead, IBM first failed in public cloud and then struggled to come up with a credible hybrid cloud story. Red Hat could give it the latter and thereby a place at the table in the former. That’s why the deal makes a lot of sense.
IBM will keep Red Hat’s open source open
Enterprises that have loved Red Hat’s open source approach don’t need to worry about that changing much under IBM’s stewardship. IBM was, after all, the company that first made open source (and specifically, Linux) safe for the enterprise.
Microsoft rightly gets credit for embracing Linux on Azure as it goes after enterprise developers, but the only reason enterprises care about Linux today is because of what IBM did way back in 2001, betting $1 billion on Linux.
No, IBM’s open source ethos isn’t quite as insistent (or, at times, dogmatic) as Red Hat’s. IBM’s more Apache than Free Software Foundation in its demeanor— more pragmatist. But, again, for enterprises that have looked to Red Hat for an open source stack, IBM is unlikely to change that openness, though there is a real risk that having spent 30 percent of its market cap on a 4 percent increase in revenues (through the Red Hat purchase), as Justin Cormack highlights, IBM may feel pressure to push Red Hat in ways unfriendly to open source developers to bring in cash.
Or not. If anything, this could give IBM a reason to more aggressively embrace open source, opening it up to a new class of customer that has hitherto steered clear of the lumbering giant.
What should developers do?
Which leaves real question looming over this deal: Will developers care? After all, it is developers who have pushed enterprise workloads into the public cloud, indifferent to the micromanagement of the CIOs sitting somewhere above them in the org chart. Those same developers are making AWS, Microsoft Azure, and Google Cloud Platform grow dramatically faster than hybrid cloud vendors like Red Hat and IBM.
Oh, and those big public cloud vendors have also partnered with VMware (or, in Microsoft’s case, have a strong hybrid story even without VMware) to deliver on hybrid cloud needs. So it’s not clear that this deal fills an unmet need for early-adopter developers.
Except, of course, that the deal isn’t about early adopters. At all.
As Drupal founder Dries Buytaert posits, “Hybrid cloud is just now setting up for growth.” Indeed, hybrid cloud really isn’t about early adopters at all. It’s about mainstream, slower-moving enterprises that want to be young and hip but aren’t, and never will be. This, by the way, describes the overwhelming majority of enterprise workloads.
As Red Hat Executive Vice President Paul Cormier puts it, “For most corporations, hybrid cloud is the only practical way to the cloud.” Perhaps more accurately stated, while pockets of enterprises have managed to jump to public cloud through their developers, the majority of applications are mired in legacy infrastructure and will need a hybrid approach to get to the cloud. For such, an IBM-Red Hat combination makes a lot of sense.
The new enterprise Linux
Ironically, this hybrid cloud emphasis has been enabled by an open source gift from one of the top cloud providers: Google’s Kubernetes. It is Kubernetes, and the containers it orchestrates, that is giving developers a new way to push workloads to the cloud. As former Hortonworks executive Shaun Connolly states, “This move signals the tipping point whereby containers (and Kubernetes) act as the layer upon which all new value is created (not Windows, not Linux). True hybrid choice is enabled by this new ‘lightweight rail.’”
In sum, not every developer can push his or her workloads to the cloud. The enterprise is simply too messy to make this manageable. Not all at once, anyway. Containers help, but containers also give enterprises a chance to take an interim step to full public cloud adoption, and that is the hybrid cloud.
By buying Red Hat, IBM just bought itself a clue in hybrid cloud computing. And it just might work.
