shivaam commented on issue #62246: URL: https://github.com/apache/airflow/issues/62246#issuecomment-3981615565
Tested multi-team ECS executor config isolation with two teams sharing the same AwsEcsExecutor class but with independent configurations. Setup: Airflow running from main (3.2.0dev) on EC2 Two teams: team_alpha and team_beta Each team has its own [team_name=aws_ecs_executor] config section pointing to separate ECS clusters and task definitions Worker images built from source using breeze prod-image build and pushed to ECR Config: ``` [core] executor = LocalExecutor;team_alpha=AwsEcsExecutor;team_beta=AwsEcsExecutor [team_alpha=aws_ecs_executor] cluster = alpha-cluster task_definition = alpha-task-def:2 [team_beta=aws_ecs_executor] cluster = beta-cluster task_definition = beta-task-def:2 ``` Results: Triggered alpha_simple_dag → ECS task launched in alpha-cluster with alpha-task-def:2 ✅ Triggered beta_simple_dag → ECS task launched in beta-cluster with beta-task-def:2 ✅ No cross-contamination: alpha tasks only appear in alpha-cluster, beta tasks only in beta-cluster ✅ Bundle routing confirmed from ECS task command payloads: - Alpha: bundle_info.name=team_alpha_dags, dag_rel_path=alpha_dag.py ✅ - Beta: bundle_info.name=team_beta_dags, dag_rel_path=beta_dag.py ✅ My tasks failed as the worker image was older but the task were triggered correctly. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
