Share this post:
The day before the opening of the OpenStack Summit in Barcelona, I already knew that one of the hot topics would be containers.
Lots of people are excited these days about containers and microservices, seeing them as the magic wand that will help IT keep up with the pace of our rapidly evolving world.
If you care about containers or platforms such as Cloud Foundry — which basically abstract you from the underlying infrastructure — you may wonder why on Earth you should care about OpenStack and a summit about it. Here’s a hint: as OpenStack COO Mark Collier mentioned in his initial keynote, the rate of adoption of container technologies is three times higher among OpenStack users than any others. That didn’t happen by accident.
Imagine you are a “container-focused” IT director. The first thing you’ll want is to deploy your container environment (Kubernetes, Swarm or Mesos). And you’ll want to do so from a console with just a click, as many times as you need, and, potentially, in different infrastructures. You can do that with OpenStack Magnum very easily. It’s not so easy without OpenStack.
Even if you want to do it on physical machines for performance reasons, you could do it — thanks to the Ironic bare metal provisioning program — exactly as in any virtualized environment.
Imagine also you need to extend your container clusters across several clouds, on premises or off premises, from one provider or several. Try doing that if each infrastructure is controlled by a different vendor tooling. You’ll wish they were all managed the same way, which is exactly what OpenStack does.
But OpenStack can provide not only a common cloud operating software to have equally managed environments, it can also ease network interconnection, too. Many of the newest features in OpenStack Newton have to do with enhancements in network connectivity.
In a nutshell, OpenStack can bring together infrastructures from different vendors because it is a real community effort. That’s its beauty and its power.
You may wonder if it is really true, tested and confirmed that OpenStacks from different providers are actually interoperable in the ways I’m describing. Indeed they are. One of the most spectacular moments of the summit was the live demo led by IBM Cloud Strategy General Manager Don Rippert, who challenged each provider to deploy the same exact application (same architecture, topology, security, and so on) in its particular OpenStack and test it. We are not talking about a couple of providers. We are talking about the major providers, nearly 20 of them, including RedHat, Huawei, Mirantis and IBM. And, believe me, it worked.
So you may be focused on containers because you want to focus on agile development. You may not care too much about the underlying infrastructure. But if you want to quickly deploy new container environments and have the ability to grow and expand your cloud to multi-cloud environments quickly, it’s wise to start caring about OpenStack.
As Don Rippert put it in his keynote: it’s not that the different vendors don’t want to compete against each other, but by collaborating in an effort like OpenStack, we are “rising the tide” to give a maximum benefit to our users, whatever their strategy.
Learn more about IBM Cloud open source technology.