Message Queues

Kafka topic/partition semantics, per-partition ordering, offsets, replication, ISR, high watermark, retention, and reassignment.

Kafka tuning as a balance among throughput, latency, durability, batching, compression, partition count, storage, and consumer lag.

How to benchmark Kafka producer, broker, and consumer capacity with realistic record sizes, partitions, replication, acknowledgements, compression, and end-to-end lag.

What managed Kafka services operate for you, what semantics remain your responsibility, and how to evaluate compatibility, networking, scaling, and cost.

A value-first route through Kafka source: protocol requests, producer batching, broker append/fetch, replication, group coordination, and KRaft metadata.

RabbitMQ fundamentals: exchanges, queues, bindings, routing, acknowledgements, publisher confirms, and modern quorum-queue high availability.

RabbitMQ exchanges, bindings, queues, direct/fanout/topic routing, work queues, acknowledgements, prefetch, and dead-letter patterns.

Publisher confirms, durable queues, persistent messages, quorum queues, consumer acknowledgements, redelivery, and why end-to-end idempotency is still required.

Modern RabbitMQ cluster responsibilities, queue data placement, quorum queues, leader/follower replication, and the legacy status of mirrored classic queues.