RocketMQ 5.0 Overview: A Cloud‑Native Messaging, Event and Stream Fusion Platform

The presentation begins with a brief history of RocketMQ, tracing its origins from Alibaba's early message engines (Notify, Napoli) to the open‑source MetaQ, the birth of RocketMQ 3.0, its graduation to an Apache top‑level project, and the rapid development of the 4.x series that introduced multi‑replica support, richer message types, and extensive community contributions.

It then outlines why a new 5.0 version is needed: the growing demand for real‑time data processing in addition to traditional business messaging, and the increasing service‑level expectations of cloud‑native customers.

RocketMQ 5.0 is positioned as a cloud‑native, hyper‑converged platform for messaging, events, and streams. Its core characteristics include high SLA and low cost, full schedulability, extensibility through an open ecosystem, automatic scaling, and adherence to industry standards.

The architecture introduces several key innovations: a lightweight SDK with gRPC support, an ultra‑minimalist design without external dependencies, a separable storage‑compute model that allows brokers to be stateless, multi‑mode storage (three‑replica Raft, two‑replica block, or single‑replica cloud‑disk), and deep integration with cloud‑native observability tools such as OpenTelemetry and Prometheus.

A new POP consumption model eliminates rebalance‑induced latency by allowing any consumer to fetch messages directly from any broker queue, using queue‑level locks and CK messages for reliable acknowledgment and retry handling. This, together with existing PUSH and PULL modes, enables a unified stream‑batch data access pattern.

Elastic scaling is achieved through logical queues that decouple physical queue placement from logical ordering, allowing second‑level cluster expansion without affecting stream ordering.

Finally, RocketMQ 5.0 will ship a lightweight real‑time compute framework called RocketMQ Streams, offering map, filter, window, join, and table operators, compatibility with Flink SQL, and user‑defined functions, while maintaining zero external dependencies.

The roadmap includes full gRPC support, additional protocols (MQTT, AMQP), further enhancements to storage‑compute separation, and continued community involvement to evolve the platform into a comprehensive messaging‑event‑streaming solution.