Interactive Kafka Internals Explorer — click anything to see it in action
0
Messages Produced
0
Messages Consumed
0
Consumer Lag
0/s
Throughput
Actions
End-to-end data flow — Consumer Group (below) coordinates but is not in the data path
⚙️
Producer
Your App / Service
ack
Sends messages via
KafkaProducer API
TCP / TLS
broker node — partitions stored here
🗄️
Broker
kafka-broker-1:9092
Receives, persists &
serves messages
appends to log
📦
Topic
orders-events
Immutable append-only
log — stored on broker
Records (Poll)
🎯
Consumer
Sink / Downstream
Fetches from broker &
commits offsets
offset commits
👥
Consumer Group
group: analytics-cg
Coordination only —
assigns partitions & offsets
How It Works
👆 Click anything to explore
Select any component in the pipeline above, or use the action buttons to simulate real Kafka behaviour. Each click shows you what happens internally — offsets advancing, partitions filling, consumer lag changing, and replication propagating.

Quick tip: Try Produce Message first, then Consume & Commit to see the full flow from source to sink.
00:00Kafka visualizer ready. Click any button or component above ↑
Topic Internals — Partitions & Offsets
orders-events
3 partitions · RF=2 · retention: 7d
__consumer_offsets
Internal topic · Kafka-managed
Kafka uses this internal compacted topic to durably store consumer group offsets. When a consumer commits, a record is written here — this replaced the earlier ZooKeeper-based offset storage in Kafka 0.9. ZooKeeper itself was replaced by KRaft for cluster metadata (broker registration, leader election) in Kafka 3.3+ — a separate concern from offset storage.
Consumer Group: analytics-cg (coordination layer — each consumer fetches directly from its broker partition leader)