You probably think that OpenStack has been supplanted by Kubernetes. Even here at Mirantis, where for many years of our existence we were known as “The OpenStack company” and where we still actively develop our own OpenStack distribution, we don’t tend to talk about it very much. But in the last year, OpenStack adoption has grown 66% to over 25 million cores, according to Mark Collier, COO of the Open Infrastructure Foundation. We haven’t seen that rapid a rise in OpenStack adoption since the early days, in the mid 2010’s. So what happened?
Simply put, with everyone working from home, companies have had to rethink their entire architecture strategy, providing the centralized infrastructure that they likely had needed for years, but just hadn’t gotten around to putting in.
Why OpenStack, and why now?
At the beginning of the pandemic, companies suddenly needed to provide infrastructure for remote workers so that they could keep functioning even with all of those empty offices. For many, it meant that they finally had to do that “digital transformation” (sorry for the buzzword) that they’d been putting off for so many years.
The thing is, all of the issues that had kept them from doing this transformation thus far were still there. They still had to ensure that they had control over their data. They still had to ensure that security was maintained at all levels of their infrastructure. They still had to solve the problem of finding people on staff (or from outside sources) that had the knowledge required to make the transformation.
Only now, they had to do it immediately, and they had to do it while watching that bottom line like a hawk because nobody knew what the global economy was going to do. It was like Ginger Rogers, partner to the extraordinary dancer Fred Astaire said: “I had to do everything Fred did, only backwards and in heels.”
For OpenStack, this has meant a double boon.
Some companies decided to keep their infrastructure in-house, and all of the reasons that already existed to switch to OpenStack were still there. They needed virtual machines but didn’t necessarily want to pay the VMware Tax, so OpenStack was a way to achieve the infrastructures they needed in an open source manner.
Other companies decided to go the public cloud route, but that’s good for OpenStack too, because OpenStack is actually the foundation of many public cloud providers, running in 170 public cloud data centers around the world.
What it comes down to is sudden demand. “Several years’ worth of future growth have been pulled in,” Collier told a press conference for the 24th release of OpenStack, Xena, “and all of a sudden, everyone needs more infrastructure.”
Not your parents’ OpenStack
Those who are surprised that OpenStack is still doing so well are likely thinking of it from its early days, when deploying and managing OpenStack was not for the faint of heart. These days, however, OpenStack is a much more manageable beast, thanks to distributions such as Mirantis OpenStack.
For example, OpenStack has now been containerized, so it can run as a Kubernetes application, as it does with Mirantis OpenStack. This capability has led OpenStack to be used for hyperscale applications much more than before.
What’s more, Collier pointed out that companies that had already been using OpenStack are using even more of it. For example, China Mobile now has over 6 million cores running OpenStack in public and private environments, and while you might expect that kind of scale from a communications service provider, that’s not where it stops; Walmart is now running OpenStack on more than 1 million cores.
What’s more, while you may think of Windows when you think of Microsoft, as the company behind Azure, they’re also the second largest public cloud provider in the world. For years the company has been contributors to OpenStack, and now they are Platinum members.
The OpenStack ecosystem is also much larger than it was in the early days, when it consisted of just a handful of projects. Now there are dozens of components you can add to beef up your OpenStack deployment, covering capabilities such as resource optimization, billing, and business logic.
What happens next?
When all of this started, news outlets were agog at the amount of time, money, and effort being spent, noting that for some companies, money was no object in this transformation, because it was absolutely essential to their survival. They wondered what was going to happen when the emergency was over. Would companies go back to their old ways?
It’s unlikely. For one thing, making the transformation to OpenStack wasn’t a frivolous act of panic. It was merely an acceleration of the transformation companies needed to make anyway. For another, OpenStack serves as a great base not just for typical day-to-day work, but also for newer technologies. (In other words, while you can run OpenStack on Kubernetes, you can also run Kubernetes on OpenStack.) Whatever new technology you’re thinking about, be it containers, no-code/low-code, or serverless, it still needs infrastructure behind it.
So while I’ve lost track of how many times I’ve had to write “No, OpenStack isn’t dead,” blogs, I’m beginning to think that this time, it’s going to stick — just like OpenStack itself. So if you want to give it a try, please go ahead and contact us.