Thanks for joining in here, and I must say the reason I want to facilitate this session is that I wasn’t all that clear on this new wave of infrastructure. That said, I was immediately struck by how much the idea of thinking about server infrastructure in terms of shipping containers piqued my imagination. The idea that shipping containers had changed the way we imagined that industry by making portability dead simple. No offloading from ship to dock via stevedores, but rather move a container full of goods from ship to truck via crane in a fraction of the time.
Now, what if we think of where and how we store our data in similar ways in terms of shipping containers. I am running a web server with WordPress, Omeka, and several other applications on my own dedicated server. To move the applications to a new server I need to manually move the files, the databases, etc. and then rebuild the environment elsewhere. With a container the idea is each of these applications would have their own container, and they would move over as a piece. You would theoretically have a small, contained server for each application that would allow you to install, run, and move this application seamlessly between servers.
That’s not all, and it might be a bit of a simplification, and I invite other to correct me, but the idea struck me. The idea that infrastructure was becoming smaller, focused and self-contained. It was moving us closer to the idea of the internet as a utility, which is, at least for me, why the idea of the cloud is fascinating. I think the best description of the cloud I ever heard was this one by Stephen Fry, which is a commercial for DataBarracks:
That helped me understand containers, because to have use-driven infrastructure, you have to be able to develop an architecture that can run at the level of the application.