In praise of boring infrastructure
Every team I have worked with that shipped reliably had one thing in common: they chose technology that nobody wanted to talk about at conferences. A few thoughts on why that works.
Read more →Retries are a distributed systems problem
A retry looks like a local decision, but it changes the load on every service downstream. Backoff, jitter and budgets, explained with a few real incidents.
Read more →Small tools, sharp edges
Why I keep a folder of forty-line scripts, and how they slowly became the most valuable code I own.
Read more →