Based on the research, I can explain what's happening and how to prevent it. The behavior you're seeing where new schedules are being added rather than replacing the old ones typically occurs when: 1. The deployment name or key parameters have changed between deployments 2. The schedule configuration isn't being properly overwritten in the deployment process To ensure schedules are properly overwritten and not accumulated, here are the key steps: 1. Maintain Consistent Deployment Names: Make sure your deployment name remains consistent across redeployments. If the name changes, it will create a new deployment instead of updating the existing one. 2. Explicit Schedule Management: In your prefect.yaml, ensure you have a clear schedule definition:

yaml
   deployments:
     - name: your-deployment-name
       schedules:
         - schedule: "0 0 * * *"  # Your schedule definition
           active: true

3. Clean Up Before Deployment: If you're seeing accumulated schedules, you can: - Delete the existing deployment before creating a new one - Use the Prefect UI to manually remove unwanted schedules - Use the Prefect API to programmatically clean up schedules 4. GitLab CI Implementation: In your GitLab CI pipeline, you might want to add a cleanup step:

yaml
   deploy_flow:
     script:
       - prefect deployment delete "flow-name/deployment-name" --skip-yes  # Optional cleanup
       - prefect deploy  # Your deployment command

5. Verify Deployment Updates: After deployment, you can verify the schedules using:

bash
   prefect deployment inspect "flow-name/deployment-name"

The most important thing is to ensure that your deployment configuration in prefect.yaml is complete and accurate, as this is the source of truth for your deployment settings, including schedules. For more details on managing schedules and deployments, you can refer to the Prefect documentation on creating schedules and deployment concepts.