RabbitMQ’s Common Issues & Misconceptions: A Support Perspective
summary
CloudAMQP support sees about 500 support issues/month, mostly caused by client prefetch misconfiguration leading to poison-message redelivery and by wrong queue types; RabbitMQ 4.0 sets max message size 16MB (configurable) and 4.1 adds a Prometheus histogram for message-size distributions, while 3.13 fixed restart rebuilds of very large messages. Operators mitigate churn and disk I/O limits with connection rate limits, proxies/pooling for PHP, and scaling out rarely improves latency; streams can hit million-messages/sec when partitioned.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:35Common Issues
- 5:58Not using optimal Q
- 7:16Connection churn
- 8:57Discs
- 10:24Misconceptions
- 10:59RabbitMQ throughput
- 12:15RabbitMQ scaling
- 13:03MQTT support
- 14:25Free metrics
- 15:28Watermark alert
- 16:28Classic mirrored cues
- 17:38Final thoughts
- 18:10Questions
related talks
30mA love letter to messaging: Celebrating RabbitMQ's journey
59mAdding 1.21 Gigawatts to Applications with RabbitMQ
32mRabbitMQ in Real Life - Tips and Tricks
22mLessons Learned from Scaling RabbitMQ in a High-Growth (...)
19mUpdate from RabbitMQ Core Team
01h 03mChoosing the Right Messaging Solution for Your Architecture
46mReal-world Examples of Message Queue Usage: Insights from RabbitMQ
27mQueue Management with RabbitMQ