Moving clouds.. with Docker

The driving force in IT for me has always been commoditization. There is probably another blog post here, but we moved from the Bios to the processor, to storage and the network. Everything becomes a commodity at some point, or do you still care what BIOS version your server has? If you still have a server?

In 2008 I attended a CloudCamp where people were asking vendors: “but how do I move workloads between cloud providers?”. Of course there was no answer because public cloud was not a commodity back then, it was just an emerging tech and they all wanted to monetize. Compare your AWS prices from back then to now if you don’t believe me…

Today I realize that containers are making cloud a commodity finally. It does not matter if you run your Docker containers in AWS or Azure or other cloud services. The apps and services remain the same. With Docker Enterprise Edition – the vCenter of Docker – you can manage them all from one pane. I realized that even more with the announcement of IBM Cloud support in Docker EE in the keynote this morning. So now you have AWS, Azure and IBM cloud support built-in for your swarm clusters (and soon kubernetes). Google cloud is experimental at the moment I understand. I played around with Docker for Mac CE deploying a swarm to AWS and I must say the experience is seemless.

So there you go: when all your apps run in containers, they can be moved around on different clouds or even run on different ones in one stack!

So where is the ‘spiel’ now you ask? In added services from those cloud providers. IBM showed in the keynote adding image recognition and Watson lookup in a traditional pet store app. Watson, Alexa and the likes are the proprietary differentiators now. For a next blog…


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.