Evolutionary History of the RabbitMQ .NET Client Towards Concurrency

Explore the evolutionary journey of the RabbitMQ .NET Client towards concurrency in this comprehensive conference talk. Dive into the client's transition from a synchronous resource-intensive system to a modern, asynchronous, and concurrent architecture. Learn about the client's background, the introduction of asynchronous code paths into a synchronous codebase, iterative improvements in message consumption loops, and the impact of async and concurrency on overall performance. Discover key concepts such as connection objects, models, confirms, and various concurrency techniques including async void, buffer copying, semaphores, and advanced interleaving calls. Examine the implementation of Consumer Work Service, CPU performance optimization, batching work pools, and prefetch count adjustments. Gain insights into benchmark results, async connection factories, task completion sources, and threading channels. This talk provides a deep understanding of the RabbitMQ .NET Client's evolution and equips developers with knowledge to leverage its concurrent capabilities effectively.

Introduction
Agenda
What is the client
What is the broker
Connection objects
Model
Confirms Aware
Demo
Why concurrency
Multiple models
How to achieve concurrency
Async void
Copy the buffer
concurrency
semaphore
advanced trickery
interleaving calls
sequential execution
concurrent exclusive scheduler
concurrency demo
buffer copy
recap
Consumer Work Service
CPU Performance
Batching Work Pool
Prefetch Count
Benchmark Results
Async Connection Factory
Task Completion Source
Threading Channel
concurrency loop
summary