About

I'm Mira, a platform engineer currently based in Vienna. I've been doing infrastructure work for about eight years — first at a fintech startup where "platform team" meant me and one other person sharing an on-call rotation, and more recently at a mid-size SaaS company where the problems are bigger but at least I'm not alone when things break.

My daily work involves Kubernetes, Terraform, Postgres, and a lot of Prometheus dashboards. I've developed opinions about all of them, some of which turn out to be wrong in interesting ways, which is usually how I get ideas for posts.

I started this site because I kept writing long internal postmortems that nobody read, and I thought: maybe if I put this on the internet someone will argue with me in a productive way. That has happened exactly twice. Both times were worth it.

What I write about

Mostly: things I had to learn the hard way. Resource limits in Kubernetes that behave differently than the docs suggest. Postgres autovacuum tuning that takes months to actually validate. On-call patterns that are supposed to reduce alert fatigue but introduce subtler problems.

I try to write about the specifics — actual numbers, actual error messages, actual timelines. Vague posts about "best practices" are everywhere. I want to write the thing I was looking for but couldn't find.

Tech I use

Kubernetes (bare metal and cloud), Postgres 14/16, Prometheus + Grafana, Terraform, Ansible for the stuff Terraform can't reach, Go for tooling. I've been experimenting with eBPF for observability; ask me about it and I'll either say something useful or just send you Brendan Gregg links.

Contact

hello [at] lisdodde.cfd — I read everything, reply to about half. No cold pitches, no sponsored posts, no link exchange requests.