Update from Rabbit Core Team
summary
RabbitMQ 3.9 ships a new streams: non-destructive log with replay and massive fan‑out via a new high‑throughput protocol (Java client mature, Go works, other clients wanted), JSON logging via Erlang logger, and cluster/peer‑discovery fixes. Classic queues see about 50% memory reduction, roughly 2x throughput and lower latency; quorum queues gain per‑message TTL and federation. Tanzu adds an OSR active-passive replication plugin for quorum queues/streams and work to replace Mnesia with capri raft metadata store for stronger partition resilience. Tests run faster via a monorepo and Bazel.
AI-written summary · about
chapters (16)
- 0:0000:44 - Introduction
- 0:4504:42 - RabbitMQ Offerings
- 4:4307:49 - Industry Themes
- 7:5009:33 - Scalability
- 9:3410:33 - Streams - Upcoming
- 10:3411:20 - Classic Queues - Work in Progress
- 11:2111:43 - Quorum Queues - Work in Progress
- 11:4413:29 - Observability
- 13:3015:37 - Resiliency
- 15:3816:25 - Development Experience
- 16:2619:17 - Tanzu RabbitMQ Updates
- 19:1820:41 - Tanzu RabbitMQ Ease of Use
- 20:4221:22 - Tanzu RabbitMQ Scalability - Coming
- 21:2322:20 - Tanzu RabbitMQ from VMs (for TAS)
- 22:2124:38 - Team Update
- 24:3930:24 - QnA
related talks
14mManagement, Monitoring and Tracing in RabbitMQ
46mFollow the RabbitMQ
31mStreaming: Real-Time Architecture for Connected Products in Automotive
41mDevOps' Swiss Army knife for the BEAM
29mRealtime web applications with streaming REST
18mWebSockets and RabbitMQ: Web-STOMP, (...)
26mApache ActiveMQ and the coming features around cloud and management
33mObservability Projects Updates