from Hacker News

I'm interested in understanding how folks configure the systems they build and the reasoning behind those choices. I am mostly curious about those using online configuration services and APIs, as opposed to process-level configuration (env vars, command line arguments and files). But if you just use those instead, that is good to know too!

Some specifics that would be great to know about:

- What service do you use (something OSS, proprietary or home grown)? - What makes it such a good fit for your use-case? - Does it have any features to ensure your configuration is valid for your application? - How do you understand what changed in your configuration during an incident? - Does it favour consistency or availability? - How do your applications access it? Does it support push, pull or both? - Do you need specific permissions or peer-review to make changes to configuration?