Hi All, Our stack: Prefect Cloud + K8s + LocalDask When trying to trigger a flow using GraphQL API (within our Python backend service), we are seeing these errors sometimes..

requests.exceptions.RetryError: HTTPSConnectionPool(host='<http://api.prefect.io|api.prefect.io>', port=443): Max retries exceeded with url: / (Caused by ResponseError('too many 502 error responses'))

And

urllib3.exceptions.MaxRetryError: HTTPSConnectionPool(host='<http://api.prefect.io|api.prefect.io>', port=443): Max retries exceeded with url: / (Caused by ReadTimeoutError("HTTPSConnectionPool(host='<http://api.prefect.io|api.prefect.io>', port=443): Read timed out. (read timeout=15)",))

Are there any API limits, throttling, etc?

Yes we do have some form of throttling but the limits are pretty high. How many tasks are you mapping over?

This error is not displayed in our flow run, instead, in a separate microservice, we use the graphQL queries to trigger a flow run, there it shows this error. I assume at that time a lot of flows might be working with many mapped tasks, hence this issue?

For Prefect Cloud specifically, those are the payload limits: https://discourse.prefect.io/t/what-is-the-api-request-payload-limit-in-prefect-cloud/316