Nice!

Reddit suggested these ones:

Observability: Archiving petabytes of log, tracing, and analytics data "in case we need it". Meanwhile, Steve keeps kubectl exec ... /bin/bash into each container to view the logs "real-time".

Data Mesh: Centralized and monolithic data management is too simple and not micro-servicey enough. So DevOps made distributed data management to justify inflating salaries.