Brokers and distributed messaging
Message brokers sit between producers and consumers in the data stack, providing durable, ordered event streams that decouple ingestion from processing and enable replay, fan-out, and exactly-once delivery semantics.
- Apache Kafka - Distributed event streaming platform, the de facto standard for high-throughput data pipelines and event-driven architectures.
- Apache Pulsar - Distributed messaging and streaming platform with multi-tenancy, geo-replication, and a decoupled storage layer.
- NATS / JetStream - Lightweight cloud-native messaging system; JetStream adds persistence, replay, and streaming semantics.
- RabbitMQ Streams - Persistent, append-only log streams for RabbitMQ, enabling high-throughput message replay and fan-out.
- Redpanda - Kafka-compatible streaming data platform written in C++, with no ZooKeeper dependency and lower latency.