To get familiar with Kafka and the new DA stack.
- Rereading the entire queue is bad (but did avoid the need to develop a data-store)
- Formatting and data should be kept apart
- Duplicates aren't the wrose thing in the world
- CI/CD is easier in YAML then PowerShell/Bash etc
- Linux images can be a pain (but windows sometimes just works)
- You can't debug on other peoples machines. You need metrics and logs that are easy to digest/access