anyone have this issue with Prefect3 , in my deplo...
# ask-community
c
anyone have this issue with Prefect3 , in my deployment prefect.yaml file, I explicitely state what subnets of my VPC to run the contianer from:
Copy code
defaults: &deploydefaults
  description: Prefect 3 deployment 
  version: 1.0.0
  entrypoint: flows/my_flow.py:flow
  work_pool:
    name: flow_work_pool
    work_queue_name: '' # Use `default`
    job_variables:
      cpu: 512
      memory: 1024
      cluster: my_cluster
      task_role_arn: arn:aws:iam::xxxxxx:role/my_task_role
      execution_role_arn: arn:aws:iam::xxxxxx:role/my_task_execution_riole
      networkConfiguration:
        awsvpcConfiguration:
          subnets:
          - subnet-abc123
          - subnet-def456
          - subnet-ghi789
          securityGroups:
          - sg-987654321
          assignPublicIp: ENABLED
      image: '{{ my_image.image }}'
  schedules: []
  tags: 
    - ""
  concurrency_limit:
The 3 subnets are public - i have 3 private subnets too. But, randomly (i guess approximately half of the time?), the container is started in one of the private subnets, and this is a problem since I don't have a NAT Gateway and it will fail. Any ideas? Thanks!! 🙏
Resolved. couldn't get it working using the toml - so instead put it in the UI config for the work-pool