I’ll cover Apache Kafka core terms, Confluent Platform extensions, and Confluent Cloud additions.

Kafka & Confluent Terminology – Complete Glossary

🔹 Core Kafka Concepts

1. Kafka Cluster
   A group of servers (called brokers) working together to store, process, and stream data.
2. Broker
   A single Kafka server that stores data and serves client requests (produce/consume).
3. Producer
   An application that sends data (messages) into Kafka topics.
4. Consumer
   An application that reads data (messages) from Kafka topics.
5. Consumer Group
   A group of consumers working together to read data from a topic. Kafka ensures each message is processed by only one consumer within the group.
6. Topic
   A named channel where producers send messages and consumers read messages (like a folder or queue).
7. Partition
   A topic is divided into slices called partitions. Messages in a partition are ordered. Partitions allow parallelism and scalability.
8. Offset
   The position of a message in a partition (like a bookmark). Consumers use offsets to track what they’ve read.
9. Record / Message
   A single unit of data in Kafka. It has:
   • Key (optional, used for partitioning/order)
   • Value (the actual payload)
   • Headers (extra metadata)
10. Log
    A partition is stored as an append-only log (new messages are always written at the end).
11. Replication
    Kafka keeps copies of partitions across multiple brokers for fault tolerance.
12. Leader & Follower
    • Leader: The main replica of a partition that handles all reads/writes.
    • Follower: Copies the leader’s data for backup.
13. ISR (In-Sync Replicas)
    A set of replicas that are fully caught up with the leader.
14. Retention Policy
    Defines how long Kafka keeps data (e.g., 7 days, forever, or until size limit).
15. Compaction
    A cleanup policy that keeps only the latest value per key, deleting older duplicates.
16. Throughput
    The rate at which Kafka processes messages (messages per second).
17. Latency
    The time it takes for a message to travel from producer → broker → consumer.

🔹 Kafka Internals

18. ZooKeeper (Legacy)
    Used in older Kafka versions to manage cluster metadata and leader election. (Being replaced by KRaft).
19. KRaft (Kafka Raft Metadata mode)
    New architecture where Kafka itself manages metadata, removing the need for ZooKeeper.
20. Controller
    A special broker responsible for managing partition leaders.
21. Rebalancing
    When consumers join/leave a group, Kafka redistributes partitions among them.
22. Coordinator
    The broker responsible for managing a consumer group.
23. ACL (Access Control List)
    Security rules defining which user/app can access which topic or resource.
24. Quotas
    Limits on how much data a client can produce/consume to prevent abuse.
25. Idempotent Producer
    Ensures no duplicate messages are produced even if retries happen.
26. Exactly-Once Semantics (EOS)
    Guarantee that messages are processed only once, even during failures.
27. Transactions
    A way to group multiple messages into an atomic unit of work.

🔹 Confluent-Specific Terms

28. Confluent Platform
    An enterprise distribution of Kafka with additional tools for management, monitoring, and integration.
29. Confluent Cloud
    A fully managed Kafka service hosted by Confluent on AWS, Azure, or GCP.
30. Schema Registry
    Stores and enforces schemas (data formats) for messages (e.g., Avro, JSON, Protobuf) to ensure compatibility.
31. kSQL / ksqlDB
    A SQL-like engine to query, process, and transform Kafka streams in real-time.
32. Kafka Connect
    A framework to move data in/out of Kafka using connectors (e.g., JDBC, S3, Elasticsearch).
33. Connector
    A plugin used with Kafka Connect to integrate Kafka with external systems.
    • Source Connector: Pulls data into Kafka.
    • Sink Connector: Pushes data out of Kafka.
34. Confluent Hub
    A marketplace of prebuilt Kafka connectors.
35. Confluent Control Center
    A GUI tool for monitoring Kafka clusters, topics, connectors, and schemas.
36. Replicator
    A Confluent tool to copy topics from one Kafka cluster to another (useful for multi-region).
37. Confluent REST Proxy
    Allows producing/consuming data using REST APIs instead of Kafka clients.
38. Confluent RBAC (Role-Based Access Control)
    Fine-grained access control for Kafka resources.
39. Confluent CLI
    A command-line tool for managing Confluent Cloud clusters, topics, and connectors.
40. Tiered Storage
    A Confluent feature that offloads older Kafka data to cheaper cloud storage (e.g., S3, GCS).
41. Cluster Linking
    A Confluent Cloud feature to link clusters across regions/clouds for data replication.
42. Confluent Cloud Metrics API
    Provides usage and performance metrics for monitoring clusters.

🔹 Stream Processing Terms

43. Kafka Streams
    A Java library for building real-time streaming applications on top of Kafka.
44. Stream
    A continuous flow of data records in Kafka.
45. Stream Processor
    An application that transforms or processes Kafka data in real-time.
46. Topology
    The workflow (graph of processors) that defines how streams are processed.
47. State Store
    Local storage used by stream processing apps to maintain state (e.g., counts, aggregations).
48. Global Store
    A replicated state store available to all stream tasks.
49. Windowing
    Grouping data by time intervals (e.g., 5-minute sales totals).

🔹 Advanced Kafka Concepts

50. Reassignment
    Moving partitions across brokers for load balancing.
51. Throttling
    Slowing down producers/consumers to avoid overwhelming the cluster.
52. Backpressure
    When consumers can’t keep up with producers, causing slowdowns.
53. Dead Letter Queue (DLQ)
    A special topic where failed or invalid messages are sent for later debugging.
54. MirrorMaker 2.0
    Kafka’s built-in tool for replicating data across clusters (open-source equivalent of Confluent Replicator).
55. Metrics & JMX
    Kafka exposes metrics via JMX for monitoring cluster health.
56. Log Segment
    Each partition’s log is broken into smaller files called log segments.
57. Message Key Partitioning
    The method Kafka uses to decide which partition a message goes to (based on key hash).
58. Rack Awareness
    Kafka spreads replicas across different racks/data centers for reliability.

Rajesh Kumar

I’m Rajesh Kumar, a DevOps, SRE, DevSecOps, Cloud, and Platform Engineering expert passionate about sharing practical knowledge, real-world experiences, and industry best practices. I have worked at Cotocus and regularly write about technology, travel, investing, health, product reviews, and digital marketing through my various platforms.

I publish technical articles at DevOps School, travel stories at Holiday Landmark, stock market insights at Stocks Mantra, health and fitness guidance at My Medic Plus, product reviews at TrueReviewNow, and SEO and digital marketing strategies at Wizbrand.