This repository is a suite of operational tools for Apache Kafka clusters, originally developed by DataDog for internal use and open-sourced. It consists of four primary binaries: `topicmappr` for partition reassignment and storage rebalancing using bin-packing algorithms with rack awareness; `registry` for a gRPC/HTTP API service that manages Kafka resource metadata (topics, brokers) with custom tagging and offset translation; `autothrottle` for dynamically adjusting replication throttle rates based on Datadog metrics to prevent network saturation; and `metricsfetcher` to pull Datadog metrics into ZooKeeper for consumption by the other tools. The architecture is modular, with shared libraries for Kafka admin operations (`kafkaadmin`), ZooKeeper interactions (`kafkazk`, `cluster/zookeeper`), metrics retrieval (`kafkametrics`), and partition mapping algorithms (`mapper`). The system relies heavily on ZooKeeper for state storage and distributed locking, with a partial migration path to Kafka native APIs (KIP-455) in the autothrottle component. The codebase demonstrates strong engineering practices including interface-based abstraction, comprehensive test stubs, integration test suites, and Docker-based reproducible environments. However, it shows signs of technical debt such as legacy code paths, hardcoded constants, and dependency on an older Datadog client library. The tools are designed for production Kafka cluster operators and assume familiarity with Kafka administration, ZooKeeper, and Datadog monitoring.