DevOps is taking over the tech world. Now, automated programmable infrastructure is reaching widespread adoption. We’ve also embraced test automation. But there’s a crucial part we’re missing: testing the programmable infrastructure code itself. After all, it’s code too, right?
Even worse, microservices and other exotic architectures are making deployments ever more complex. The more complex things get, the more bugs we will find. We can’t afford to ignore it for long.
I worked on a project where we absolutely had to test infrastructure - because our application deployed, configured, and maintained cloud resources. This talk will take you through some of the strategies we developed, some of the problems we encountered, and some thoughts about where the industry might be headed from here.
- Why infrastructure also needs testing
- Why testers need to care about infrastructure and why it can't just be left to ops
- What kind of tools are available to test it
- What approaches you can take to test it
- What downsides are there to it, and what problems we face now