If you are planning to move to a micro-services architecture from a monolith, one of the biggest changes you will face is the change to your deployment process. Monoliths traditionally have a significant end-to-end test suite that needs to run before a release decision can be made. If your customers depend on a low release frequency, jumping to a set of micro-services, each having its own release schedule, may not be desirable. In fact, you may not want to lose that suite of end-to-end tests.

So how do you continue to thoroughly test your product so that a release decision can be made? In this talk, I will tell you about the lessons we’ve learned at Salesforce as we’ve started making the transition to micro-services. I will cover the three types of tests you need for a micro-service architecture (base, performance/scalability, and resilience tests) and the patterns and supporting technologies (circuit breakers, distributed trace, chaos engineering) you need.

Using these techniques, you’ll not only increase your confidence in your release process, but you’ll also be able to find the root cause of problems faster, predict scalability requirements, and identify and fix production problems before they impact customers.