Writing

39 long-form articles on production engineering — how real systems fail, why, and what actually fixed them.

2 articles matching “Messaging & Queues”

Production IncidentsMay 9, 202625 min read

“We Have Exactly-Once Delivery” — The Lie Your Architecture Is Telling You

Why “our queue guarantees exactly-once” is almost always a lie once your flow crosses a component boundary: Kafka EOS scope, ack-after-process, unique constraints, SQS FIFO, DB transactions with SMTP, retries without…

Kafka & PipelinesSep 25, 202625 min read

How Kafka Finally Solved the “One Slow Message Blocks Everything” Problem — Share Groups, Explained Simply

You add 24 more workers to fix Kafka lag — and they all sit idle. Meanwhile one huge message is stuck at the front of a partition, and nothing behind it moves.