My workplace uses a lot of serverless features on AWS - Lambdas, SWF, SQS, API gateway, as well as auto scaling instances on ec2/eb. Some of it’s nice. Most of

My workplace uses a lot of serverless features on AWS - Lambdas, SWF, SQS, API gateway, as well as auto scaling instances on ec2/eb. Some of it’s nice. Most of it isn’t. Here are the pros: it’s great for quickly scaling up or down, and we don’t have to worry about physical infrastructure. Here are the cons: Documentation is inadequate. The docs are often verbose, but outdated or incomplete in critical ways. There also seem to be very few resources/blog posts about how to use big aws features. I suspect there is a secret cabal of infrastructure engineers that hoard this information. Replication of work is difficult. So much is done through the GUI. This seems really amateurish. How do you quickly capture the state of your entire aws configuration? How do you record and track changes? We have very little insights into our system. Every microservice is a snowflake. Why do I have to use this apache templating language for api gateway? Weird. Who designed these libraries? Why can’t I rip as much as I want out of a queue? Weird.