A message from our sponsor, Dragonfly:
Redis runs one core per process. Your server has 64. Dragonfly is a Redis/Valkey-compatible data store, rebuilt multi-threaded by an ex-AWS ElastiCache engineer:
- 6M ops/sec on one node, no cluster
- Drop-in: same clients, same commands
- Instacart migrated and cut latency by 50% using 70% fewer nodes
Events like the recent compromise of CA signing keys are discussed out in the open, on Mozilla’s Bugzilla. Right now, you can watch the post-incident analysis as it unfolds.
Mozilla
This article gives a really easy-to-follow explanation of the consistency issues you get when writing across two systems and how to solve them with an outbox pattern.
Rahul Tewari — DZone
An explanation of why a service catalog is so important, especially during incidents.
Twenty minutes in, you haven’t started diagnosing anything. You’ve been doing archaeology.
Bonus: part two on keeping your service catalog up-to-date.
Bill Duncan
The Firebase SDK caused app-side crashes due to a server-side data change. This article focuses on Google’s response, and I also recommend reading the writeup on the Firebase blog that Gergely links to.
Gergely Orosz — The Pragmatic Engineer
LLM agents behave fundamentally differently from deterministic programs, and we can’t just expect them to work like a natural language programming language.
Uwe Friedrichsen
This one delves into the design choices involved in running tons of little databases at the edge efficiently.
David Laban and Lambros Petrou — Cloudflare
A seasoned emergency responder and computer incident management engineer shares a rare perspective on the similarities and differences between the two fields.
Brent Chapman
Now I start a chat session with an AI model, give him a personality, most often it is an old and grumpy but well-mannered distinguished engineer…
Hrvoje Rancic — ShiftMag