Error with ECS container: ResourceInitializationError : failed to create Cloudwatch log stream

Question

I am trying to set up an ECS Fargate container but it throws me this error:

"ResourceInitializationError: failed to validate logger args: create stream has been retried 1 times: failed to create Cloudwatch log stream: ResourceNotFoundException: The specified log group does not exist. : exit status 1"

I've already checked and the log group exists and it has the same name in the task definition. I've checked the ecsTaskExecutionRole policy (it has cloudWactchLogsFullAccess policy), also thought that could be the internet access but I checked the vpc, subnet, and route table and everything seems ok (I don't know how to check if the container really has internet access).

Answer 1

One of my Fargate services had mysteriously stopped and I was seeing this error as well.

In my case, I had deleted the log group which the service was using and this caused the launching of the task to fail on subsequent retries.

You say that your log group still exists though, so I'm not sure if it's the same problem.

Anyway, my FIX using the AWS Console:

1. Navigate to the task that failed to launch
2. View in CloudWatch
3. Get the name of the CloudWatch log group that the task is expecting to save to
4. Create a new CloudWatch log group matching with this name

Now the service has a log group to save to and your tasks should successfully start again.