For many years, the contact point between development and operations has been always a source of problems when deploying a new version of an application to production. Different languages generate different types of artifacts (war or JAR for Java, the source code for Node.js.), which led to heterogeneity in the procedures when rolling out new versions.

This heterogeneity led into bespoke solutions to roll out versions, which are pretty much sorceries with weird habits, such as deploying at 4 a.m. to avoid an outage in the system and creating error-prone bash scripts that are harder to maintain than the software itself. The problem, aside from the complexity, is that new hires need to ramp up into your systems, and this always introduces a level of risk that we are not aware of for the majority of the time until something goes very wrong.

Docker came to the rescue. With Docker, we can generate a deployable artifact, which is not only the software that you built but also its...