@Marvin How not to stop running remaining tasks while previous parallel tasks failed?

To ensure remaining tasks in a Prefect 3.x flow continue to run even if some parallel tasks fail, you can take advantage of Prefect's error-handling features. Prefect allows you to set policies and behaviors that help control how tasks respond to failures. 1. Task Retry Policies: Prefect tasks can be configured with custom retry behavior using options such as

retry_delay_seconds

and

retry_condition_fn

. These allow you to specify how often and when retries should occur if a task fails. You can also specify different retry intervals and conditions for specific error scenarios. See more here. 2. Failure Handling with ControlFlow: Prefect's ControlFlow extension allows for advanced failure handling, such as specifying conditions under which a flow or orchestration should terminate or continue. By using "run_until" conditions, you can set complex logic that determines when the flow should stop based on task failures, completions, or custom criteria. Read more about it here. By setting appropriate retry policies and using advanced control logic, you can ensure that your tasks are resilient and continue running despite individual failures. Just another challenge for us, mere mortals, to solve!