DevOps
Rolling, blue-green or canary? Choosing a deploy strategy
Every deploy strategy trades speed, cost and blast radius differently. Here's how the three common ones compare and when each makes sense.
Tag
#reliability
Every story tagged #reliability.
DevOps
Health checks that mean something: liveness vs. readiness
A health check that always returns 200 is worse than none. Separate "is the process alive?" from "can it serve traffic?" and your deploys get safer.
Web
Background jobs that don't bite: idempotency, retries and timeouts
Jobs will run twice, run late and fail halfway. Designing for that from the start is what separates a reliable queue from a source of 3 a.m. pages.
Security
Rate limiting algorithms: token bucket, fixed window and sliding window
Rate limits protect logins from brute force and APIs from overload. The algorithm you choose decides how bursty traffic is treated.
DevOps
SLOs and error budgets: deciding how reliable is reliable enough
100% uptime is the wrong goal. Service level objectives turn reliability into a number you can plan around, and error budgets tell you when to slow down.