Every Apache Kafka command you actually need — wrapped in a Makefile so you stop googling flags.
Running Kafka day to day means retyping the same twelve commands with slightly different flags.
This repo turns them into one-word make targets, ships production-shaped config files, and keeps a
full command reference for the moments you need the raw CLI.
Built from production experience operating Kafka platforms at scale. More guides on tuning, security and operations: floriancourouge.com/blog
Kafka 4.0 removed ZooKeeper entirely. Clusters run on KRaft, and the bootstrap sequence changed: you now format the storage directory with a cluster ID before the first start.
# 1. Download and extract (Java 17+ required)
wget https://downloads.apache.org/kafka/4.3.1/kafka_2.13-4.3.1.tgz
tar -xzf kafka_2.13-4.3.1.tgz
cd kafka_2.13-4.3.1
# 2. Generate a cluster ID
KAFKA_CLUSTER_ID="$(bin/kafka-storage.sh random-uuid)"
# 3. Format the log directory (once per node, before the first start)
bin/kafka-storage.sh format --standalone -t "$KAFKA_CLUSTER_ID" -c config/server.properties
# 4. Start the broker
bin/kafka-server-start.sh config/server.propertiesThen point this repo at it:
make check # verify Java, Kafka and Python are in place
export KAFKA_HOME=/path/to/kafka_2.13-4.3.1
export KAFKA_SERVER=localhost:9092
make topic-create TOPIC=ordersStill on Kafka 3.x with ZooKeeper? Read this before upgrading.
You cannot jump straight from a ZooKeeper-based cluster to 4.x. The supported path is:
- Upgrade to a recent 3.x release (3.9 is the last one with ZooKeeper support).
- Migrate that cluster to KRaft mode while still on 3.x.
- Only then upgrade to 4.x.
The --zookeeper CLI flag disappeared well before 4.0 — every command in this repo uses
--bootstrap-server, which works on 3.x and 4.x alike.
The Makefile is the point of this repo. make help lists everything; here are the ones you'll use daily.
| Area | Command | What it does |
|---|---|---|
| Setup | make install |
Install the Python dependencies |
make check |
Verify Java, Kafka and Python prerequisites | |
| Broker | make start / make stop / make restart |
Control the local broker |
make status |
Is it up, and on which port | |
make logs |
Tail the broker logs | |
| Topics | make topic-create TOPIC=orders |
Create a topic |
make topic-list |
List topics | |
make topic-describe TOPIC=orders |
Partitions, replicas, ISR | |
make topic-delete TOPIC=orders |
Delete a topic | |
| Consumer groups | make groups |
List every group |
make group GROUP=my-app |
Lag, offsets and members for one group | |
make reset GROUP=my-app TOPIC=orders |
Reset offsets | |
| Benchmarks | make test-producer / make test-consumer |
Throughput tests |
make test-batch / make test-compression |
Compare batching and codecs | |
make test-full |
Full performance suite | |
| Analysis | make analyze-health |
Under-replicated and offline partitions |
make analyze-offsets |
Offset and lag breakdown | |
make analyze-size |
Topic sizes on disk | |
make analyze-full |
Everything above in one run | |
| Examples | make producer / make consumer |
Run the Python examples |
| Cleanup | make clean / make test-cleanup |
Remove generated and benchmark data |
The four scripts behind these targets also run standalone. They look for the Kafka binaries in
$KAFKA_HOME/bin first, then on the PATH, and accept both naming conventions
(kafka-topics.sh from the Apache tarballs, kafka-topics from the Confluent packages).
Both KAFKA_HOME (default /opt/kafka) and KAFKA_SERVER (default localhost:9092) are
overridable, so the same targets work against a remote cluster:
make topic-list KAFKA_SERVER=broker-1.prod:9092├── Makefile # the interface — start here (make help)
├── COMMANDES.md # full CLI reference, every command with examples
├── docs/ # 12 chapters: concepts, security, stream processing, use cases
├── requirements.txt
└── kafka/
├── configs/ # producer, consumer, and SASL/PLAIN + SASL/SCRAM admin clients
├── scripts/ # quick-start, topic-manager, performance-tests, topic-analyzer
└── examples/ # producer-example.py, consumer-example.py
COMMANDES.md is the reference sheet: topics, consumer groups, offsets, ACLs, performance testing, troubleshooting — with the flags spelled out.
docs/ goes wider (in French): core concepts, Confluent and Conduktor, security and governance, advanced stream processing, SaaS options, and real-world use cases.
- You operate Kafka and want the daily commands one keystroke away.
- You're learning Kafka and want config files that resemble production, not a toy.
- You're benchmarking and want repeatable producer/consumer tests.
Issues and pull requests are welcome — especially additional commands, fixes for newer Kafka releases, and translations.
MIT — use it freely.
Florian Courouge — Apache Kafka & Confluent expert, DevOps engineer. floriancourouge.com · LinkedIn · Kafkabubble (multi-tenant Kafka control plane)
