Migration from On-Prem Messaging System to The Cloud
summary
Compares three migration paths: lift-and-shift to EC2, re-platform to Amazon MQ and cloud-native serverless with SQS/SNS, showing tradeoffs in operational overhead, protocol changes (AMQP/Pika vs HTTPS APIs) and scaling. Demonstrates Python examples for RabbitMQ (Pika) and SQS, explains single-node vs cluster HA with messages mirrored across three AZs, and warns of SQS limits (default 1 MB -> 2 GB via S3 pointers). Recommends refactoring producers/consumers and choosing FIFO vs standard queues.
AI-written summary · about
chapters (15)
- 0:00Intro to Messaging Migrations
- 0:44Messaging Systems Overview
- 1:24Industry Use Cases and Needs
- 2:41Why Move Off On-Prem
- 3:35Migration Paths Explained
- 4:42Shared Responsibility Shift
- 6:42Serverless Messaging Benefits
- 7:50Rehost vs Managed Brokers
- 10:55Amazon MQ Deep Dive
- 12:01Single Node vs Cluster HA
- 13:17Python Client for Brokers
- 15:16Refactoring for Serverless
- 17:31SQS and SNS Python Examples
- 19:26Key Considerations and Limits
- 23:30Wrap Up and Takeaways
related talks
39m101 ways to break your RabbitMQ
22mLessons Learned from Scaling RabbitMQ in a High-Growth (...)
48mA Beginner's Guide to Inter-Service Messaging
39mHow to Choose the Right Messaging Service for Your Workload
40mScaling RabbitMQ
50mMigrating your messaging workloads to the Cloud with Azure Service Bus
16mBuilding scalable IoT systems with MQTT and AWS
32mLeading Without the Bloat: Lessons from B...