front matter
I first tried building applications with microservices around 2013. That was the year Docker was initially released, but back then, I hadn’t heard about it. At that time, we built an application with each microservice running on a separate virtual machine. As you might expect, that was an expensive way to run microservices.
Because of the high running costs, we then opted to create fewer rather than more microservices, pushing more and more functionality into the existing microservices to the point where we couldn’t really call them microservices anymore. It was still a distributed application, of course, and it worked well enough, but the services weren’t as micro-sized as we had hoped.
I already knew at that stage that microservices were a powerful idea, if only they were cheaper. I put microservices back on the shelf but made a note that I should look at them again later.
Over the years, I watched from the sidelines as the tools and technology around microservices developed, powered by the rise (and rise) of open source coding. And I looked on as the cost of cloud computing continued to drop, spurred on by competition between vendors. Over time, it was clear that building and running a distributed application with micro-sized components was becoming more cost effective.