Topic track · 5 notes
Kafka & Messaging
Partitions, consumer groups, delivery semantics, retries and event-contract evolution.
Kafka Throughput, Partitions and Backpressure: A Practical Mental Model
How partitions, consumer concurrency, processing latency and downstream capacity determine real event throughput.
Read article ↗Kafka Consumer Groups and Rebalancing Without Surprises
How partition assignment changes, why long processing causes churn and how cooperative strategies reduce disruption.
Read article ↗At-Least-Once Delivery and Idempotent Consumers
Why duplicate delivery is normal and how deduplication, state transitions and database constraints create one durable outcome.
Read article ↗Kafka Retries and Dead-Letter Topics Without Losing Context
A failure-handling design for transient errors, poison records, backoff, replay and operational ownership.
Read article ↗Event Schema Evolution Without Breaking Consumers
Compatibility rules, additive change, semantic versioning and consumer-driven rollout for long-lived events.
Read article ↗