finds.dev

yahoo/CMAK

CMAK is a tool for managing Apache Kafka clusters

View on GitHub

★ 11,925ScalaApache-2.0updated Aug 2023

// the take

CMAK is a web UI for managing Apache Kafka clusters — adding topics, running partition reassignments, viewing consumer lag, and inspecting broker state. It was the de facto Kafka admin UI for years when Kafka's own tooling was all command-line. Last meaningfully updated in 2023 and only supports Kafka through 0.11/2.x era APIs.

Akka actor model is used throughout for concurrent broker polling, which is the right approach for fan-out state collection across many brokers. JMX polling for broker and topic metrics is genuinely useful and not something most alternatives bother with. Multi-cluster support in a single deployment is well-implemented — you can manage prod and staging from one instance. The partition reassignment UI saves a lot of pain compared to the JSON-file-based CLI workflow.

Kafka has shipped its own web UI (Kafka UI) and the KRaft mode removes ZooKeeper entirely — CMAK still depends on ZooKeeper for its own state storage, which means it's dead weight for any cluster running Kafka 3.x+ without ZK. The last commit was August 2023 and the project shows no signs of KRaft support being planned. Consumer group monitoring only covers legacy ZK-based consumers properly; the Kafka-native offset tracking via __consumer_offsets has a known single-threaded bottleneck in KafkaManagedOffsetCache that will fall behind on busy clusters. No RBAC — it's all-or-nothing access, which rules it out for any multi-team environment.

We read the README, the directory tree and a key source file in June 2026. Nobody pays to be listed. How we judge →

// at a glance

Stars
11,925
Forks
2,480
Language
Scala
License
Apache-2.0
Last commit
2 Aug 2023

// nearby

If CMAK isn't quite it.

All repos →