How do you guarantee throughput performance while facing back pressure in a push-based messaging system with failing downstream destinations? e.g., webhooks
This subject can hardly be discussed in a comprehensive way on a board like this. It also depends on what tools you have available to you - some cloud tools make this easier than building your own.
There's a great book that covers this in great detail - Feedback Control for Computer Systems: https://www.oreilly.com/library/view/feedback-control-for/97.... I've implemented many of the tactics here with great success.