AWS logging pipeline

The problem space Back in the day, a logging pipeline was a pretty manual thing to setup and manage. Yes, configuration management tools like Chef/Puppet made this somewhat easier, but you still had to run your own ELK stack (OK, it didn’t have to be ELK, but it probably was, right?) and somehow get logs in there in a robust way. You’d probably be using some kind of buffer for the logs between source and your ELK stack.