Python - Pulling ECS Service logs for a task/container that is terminating

Question

I have a python script which is grabbing logs for a ECS cluster that has multiple services. It works for tasks that are running without any issues. I have tasks in  few services that are terminating due to errors in the container scripts. I can see the logs/erros when I go to the ECS Cluster -> Click on Service and go to the Logs tab. How do I modify the script to grab those logs thrown by tasks that have terminated

My python script has the following logic.

```
response = ecs_client.list_services(cluster=cluster_name)
for service in response['serviceArns']:
  serviceResponse = ecs_client.describe_services(cluster=cluster_name, services=[service])
  serviceName = serviceResponse['services'][0]['serviceName']
  taskResponse = ecs_client.list_tasks(
            cluster=cluster_name,
            serviceName=serviceName
  )

if 'taskArns' not in taskResponse or not taskResponse['taskArns']:
            print(f"No tasks found for {serviceName}")
```

Since the task is terminating, the script always ends up saying "No tasks found.."

Accepted Answer

By default it looks for running tasks. You need to look for stopped tasks.

```
taskResponse = ecs_client.list_tasks(
            cluster=cluster_name,
            serviceName=serviceName,
            desiredStatus='STOPPED'
 )
```

desiredStatus (string) –
The task desired status to use when filtering the ListTasks results. Specifying a desiredStatus of STOPPED limits the results to tasks that Amazon ECS has set the desired status to STOPPED. This can be useful for debugging tasks that aren’t starting properly or have died or finished. The default status filter is RUNNING, which shows tasks that Amazon ECS has set the desired status to RUNNING.

Python - Pulling ECS Service logs for a task/container that is terminating

관련 콘텐츠