@phpeterle Blog Pedro Peterle
- Why we replaced a Node.js process built on change streams with Flink jobs that index into Elasticsearch and publish events to Kafka, and what changed in practice
- How Flink works inside, where it beats a hand-written consumer, and what it charges you in operations, memory and language
- What changes when you write a parallel algorithm in Bend 2, what automatic parallelism costs, and when the GPU pays off
- The Liquid Network attack, Dan Ariely's fudge factor, and why we steal just enough to keep thinking of ourselves as honest
- Traceability chains, error accumulation and calibration applied to choosing a gateway and to the reliability of a payments system
- How to build a self-hosted observability stack with OpenTelemetry, Grafana, Loki, Tempo and Mimir
- Understanding Prometheus Counter, Gauge, Histogram and Summary, with practical examples of when to use each type
-
-
-