Hi, I’m working on a project in Google Cloud Platform (GCP) with several Cloud Functions set up for ETL tasks—one for extraction, one for transformation, and one for loading. I’ve implemented these functions in an

etl.py

file, where I use

@task

and

@flow

decorators to call the deployed function URLs. This setup works perfectly when triggered manually. However, I’m encountering issues when trying to automate this with scheduled deployments. In my

deploy-etl.py

file, I specify the GitHub repository, the entry point (pointing to the ETL process in

etl.py

), the work pool name, the GCP project ID, necessary packages, and a scheduler with a cron expression. The problem is that while the deployment initially runs as expected, subsequent scheduled runs consistently show late status and a continuously loading page in the logs. The scheduled task never actually starts, despite appearing to be in progress. Does anyone have insights into why this might be happening or suggestions for troubleshooting this issue?

hi @Jose Salerno > scheduled runs consistently show late status and a continuously loading page in the logs this usually happens for a couple main reasons • the worker died somehow, so no one is around to pick up scheduled runs -> they become late ◦ can you read the logs where you ran

prefect worker start

? • concurrency slots are filled (have you set any concurrency limits?) ◦ if you go to the runs page, look for

Running

or

Cancelling

runs of your deployment that may be hanging around