Fourteen guides, grouped by when you need them. Every code sample compiles against the current release; the protocol and configuration references are checked against the crate in CI.
If you are starting from nothing, read Getting Started and then the guide for whichever client you need. If you are debugging something specific, the Configuration and Error Handling references are the fastest routes in.
Install krafka and produce and consume your first records in under five minutes.
Task-oriented recipes: the shortest correct way to do the things people actually build.
Every producer, consumer, admin and transport option, with defaults and the reason each default was chosen.
Batching, compression, partitioning, idempotence and exactly-once transactions.
Consumer groups, cooperative and eager rebalancing, offset management and KIP-848.
Queue-like consumption with KIP-932 share groups: per-record acknowledgement without partition ownership.
Topics, partitions, configs, ACLs, quotas, consumer groups, delegation tokens and cluster features.
Built-in counters, gauges and histograms, and how to export them to Prometheus.
What krafka optimises by default, which knobs actually move throughput, and how to measure honestly.
Test clients against an in-process Kafka broker with fault injection — leader moves, coordinator failover and corrupt batches, without Docker.
Error taxonomy, which errors are retriable, and how to build a dead-letter path.
