@Marvin I am seeing 429s from our prefect server deployed with 3.x and self hosted using cloud run. The 429s cause running flows to fail. Is there a way to configure rate limits for self hosted prefect servers?

It sounds like you're dealing with rate limiting issues on your self-hosted Prefect server, which result in 429 errors (indicating too many requests). With Prefect 3.x, you can manage concurrency and rate limits using Prefect’s provided utilities, which allow for controlling how many tasks, flows, or other operations can run simultaneously. Global concurrency limits are useful for optimizing resource usage and controlling task execution, while rate limits ensure system stability by governing the frequency of requests or operations. A key point is that you can create, read, edit, and delete these concurrency limits through the Prefect UI, CLI, or Python SDK. You might want to implement or adjust these rate limits to prevent excessive requests that lead to 429 responses. Here are some options you can explore: 1. Global Concurrency Limits: Use this for managing execution efficiently and preventing bottlenecks. 2. Rate Limits: Focus on system stability and fair access, suitable for preventing overuse. For detailed guidance, you can refer to the documentation on Global Concurrency and Rate Limits. Additionally, since self-hosted configurations can vary, ensure your infrastructure appropriately balances workload or utilize additional horizontal scaling to help manage load more effectively. If you encounter ongoing issues, consider checking existing GitHub Issues for any possible related discussions or solutions posted by other users. Let me know if you need further help!