@Marvin I have a GCP Cloud Run V2:push work pool. I’ve set up all my deployments to use this work pool. If a flow is already running on this work pool, and another one starts, it crashes with the following error

Flow run could not be submitted to infrastructure: <HttpError 409 when requesting <https://run.googleapis.com/v2/projects/><project>/locations/<location>/jobs?jobId=prefect-prod-cloud-run-pool&alt=json returned "Resource 'prefect-prod-cloud-run-pool' already exists.". Details: "Resource 'prefect-prod-cloud-run-pool' already exists.">

Concurrency Limit is set to Unlimited on the work pool - am i missing something?

The error you're encountering with your GCP Cloud Run V2 work pool seems to be related to a resource conflict. Specifically, it appears that Prefect is attempting to create a job with the same identifier when another flow is already running, leading to an

HttpError 409

. This error indicates that the resource already exists, which is not resolved by simply having an unlimited concurrency setting. Here are a few steps to consider: 1. Resource Naming: Ensure that each flow run gets a unique job identifier. Prefect may be trying to create a job with an existing name, causing the conflict. You might need to customize the job template or identifier to ensure uniqueness. 2. Concurrency Settings: While you mentioned that the concurrency limit is set to unlimited, it might be worth reviewing how Prefect handles job creation and whether there are any implicit limits or settings affecting job names or identifiers. 3. Check for Known Issues: There might be known issues or updates related to this. You can track such issues on Prefect's GitHub repository. Unfortunately, the search didn't return a direct match, but keeping an eye on known issues might help. 4. Documentation: Review the work pool documentation and GCP worker guide for any additional configuration details that might be relevant. If these steps don't resolve the issue, consider reaching out to Prefect support or community forums for further assistance.

@Marvin How can i make sure each flow gets a unique job identifier?

To ensure each flow run gets a unique job identifier in Prefect, you can customize the job template in your deployment configuration. This can typically be achieved by using the

infra_overrides

attribute in your

prefect.yaml

file, which allows you to override default values in the work pool's base job template. Here's a general approach: 1. Customize the Job Template: Use the

infra_overrides

attribute to specify a unique identifier for each job. This could involve using environment variables, flow parameters, or other unique identifiers available at runtime. 2. Use Dynamic Values: Incorporate dynamic values such as timestamps, UUIDs, or flow run IDs into the job identifier to ensure uniqueness. For more detailed instructions, you can refer to the Prefect YAML documentation which explains how to use deployment mechanics and overrides. If you need specific examples or further customization, you might want to explore Prefect's GitHub repository or community forums for shared solutions.

@Dominik Wagner in the work pool settings you need to keep this field blank so that prefect generates a random identifier

Ahh, thank you, @Johan sh!

@Dominik Wagner Were you able to get Google Cloud Run setup? I'm having a difficult time over here

Hey @Court, yes I was able to get it working after Jo’s suggestion - where are you getting stuck?

Good to hear that GCR works for you! @Dominik Wagner My deployments to Google Artifacts using docker are empty and I receive type various of errors including: "ImportError: cannot import name 'Infrastructure' from 'prefect.infrastructure'" ImportError("cannot import name 'Infrastructure' from 'prefect.infrastructure' (/usr/local/lib/python3.11/site-packages/prefect/infrastructure/__init__.py) The import error is the most recent error. "FileNotFoundError: [Errno 2] No such file or directory: '/opt/prefect/" I read this (https://github.com/PrefectHQ/prefect/issues/10285#issuecomment-2336734132) but still receive that issue when I use the prefect.yaml file to build my docker image from time to time. I'm on version 3.2.7 now. Are you using Dockerhub for your registry to Google Cloud Run?