remove ECS Operator retry mechanism on task failed to start

providers/amazon/src/airflow/providers/amazon/aws/hooks/ecs.py

@@ -19,7 +19,7 @@
 
 from typing import TYPE_CHECKING, Protocol, runtime_checkable
 
-from airflow.providers.amazon.aws.exceptions import EcsOperatorError, EcsTaskFailToStart
+from airflow.providers.amazon.aws.exceptions import EcsOperatorError
 from airflow.providers.amazon.aws.hooks.base_aws import AwsGenericHook
 from airflow.providers.amazon.aws.utils import _StringCompareEnum
 
@@ -38,16 +38,6 @@ def should_retry(exception: Exception):
     return False
 
 
-def should_retry_eni(exception: Exception):
-    """Check if exception is related to ENI (Elastic Network Interfaces)."""
-    if isinstance(exception, EcsTaskFailToStart):
-        return any(
-            eni_reason in exception.message
-            for eni_reason in ["network interface provisioning", "ResourceInitializationError"]
-        )
-    return False
-
-
 class EcsClusterStates(_StringCompareEnum):
     """Contains the possible State values of an ECS Cluster."""
 

providers/amazon/src/airflow/providers/amazon/aws/operators/ecs.py

@@ -26,8 +26,7 @@
 from airflow.configuration import conf
 from airflow.exceptions import AirflowException
 from airflow.providers.amazon.aws.exceptions import EcsOperatorError, EcsTaskFailToStart
-from airflow.providers.amazon.aws.hooks.base_aws import AwsBaseHook
-from airflow.providers.amazon.aws.hooks.ecs import EcsClusterStates, EcsHook, should_retry_eni
+from airflow.providers.amazon.aws.hooks.ecs import EcsClusterStates, EcsHook
 from airflow.providers.amazon.aws.hooks.logs import AwsLogsHook
 from airflow.providers.amazon.aws.operators.base_aws import AwsBaseOperator
 from airflow.providers.amazon.aws.triggers.ecs import (
@@ -688,7 +687,6 @@ def _get_task_log_fetcher(self) -> AwsTaskLogFetcher:
             logger=self.log,
         )
 
-    @AwsBaseHook.retry(should_retry_eni)
     def _check_success_task(self) -> None:
         if not self.client or not self.arn:
             return
@@ -701,11 +699,6 @@ def _check_success_task(self) -> None:
 
         for task in response["tasks"]:
             if task.get("stopCode", "") == "TaskFailedToStart":
-                # Reset task arn here otherwise the retry run will not start
-                # a new task but keep polling the old dead one
-                # I'm not resetting it for other exceptions here because
-                # EcsTaskFailToStart is the only exception that's being retried at the moment
-                self.arn = None
                 raise EcsTaskFailToStart(f"The task failed to start due to: {task.get('stoppedReason', '')}")
 
             # This is a `stoppedReason` that indicates a task has not

providers/amazon/tests/unit/amazon/aws/hooks/test_ecs.py

@@ -20,8 +20,8 @@
 
 import pytest
 
-from airflow.providers.amazon.aws.exceptions import EcsOperatorError, EcsTaskFailToStart
-from airflow.providers.amazon.aws.hooks.ecs import EcsHook, should_retry, should_retry_eni
+from airflow.providers.amazon.aws.exceptions import EcsOperatorError
+from airflow.providers.amazon.aws.hooks.ecs import EcsHook, should_retry
 
 DEFAULT_CONN_ID: str = "aws_default"
 REGION: str = "us-east-1"
@@ -59,22 +59,3 @@ def test_return_true_on_valid_reason(self):
 
     def test_return_false_on_invalid_reason(self):
         assert not should_retry(EcsOperatorError([{"reason": "CLUSTER_NOT_FOUND"}], "Foo"))
-
-
-class TestShouldRetryEni:
-    def test_return_true_on_valid_reason(self):
-        assert should_retry_eni(
-            EcsTaskFailToStart(
-                "The task failed to start due to: "
-                "Timeout waiting for network interface provisioning to complete."
-            )
-        )
-
-    def test_return_false_on_invalid_reason(self):
-        assert not should_retry_eni(
-            EcsTaskFailToStart(
-                "The task failed to start due to: "
-                "CannotPullContainerError: "
-                "ref pull has been retried 5 time(s): failed to resolve reference"
-            )
-        )