If you have more than a few microservices, their interdependencies get hard to reason about. I worked for a few years on a project to manage and control microservice dependencies and to make sure it would be possible to turn the company’s tech stack back on if we ever turned it off.
I also wrote a talk about it called “Have you tried turning it off and turning it on again?”: https://www.youtube.com/watch?v=xvk4RFGsrWo
USENIX review of the talk: https://www.usenix.org/blog/review-disaster-recovery
Graphviz examples: https://github.com/whereistanya/graphviz
And O’Reilly interviewed me about it: https://www.oreilly.com/ideas/creating-better-disaster-recovery-plans