About

OOMKilled is a field notebook from the DevOps / cloud / SRE trenches — written by a freelance consultant who spends most weeks inside other people's production systems, watching the same failure modes show up in new costumes.

The posts here are case studies and hot takes drawn from real engagements: outages, cost blowups, architecture decisions that aged badly, and the boring fixes that actually held. Details are sanitized; the lessons are not.

If your team needs help with cloud architecture, reliability, or untangling a platform that has quietly become load-bearing, that is the day job — reach out.