- Más nuevo
- Más votos
- Más comentarios
Hello,
Based on the error message you provided, it looks like there was a temporary issue with the Amazon ECS service when you tried to create your Fargate cluster called JoshTestCluster. The 500 status code in the error ("Status Code: 500") indicates that the ECS service was unavailable at the time of your create cluster request.
Let me know if u need more help regarding this.
Thanks
Abhinav
Afraid not
The role existed already, so I tried deleting and recreating it and the error is still there. I also deleted it again and let it auto create with the cluster and this did not work either (although the role was created no problems)
Can you check the CloudTrail and send it here the request parameters for the CreateCluster API call? Please remove any confidential information
That is what I thought, but I have now tried this in multiple regions over multiple days with the same error. I cannot believe that ECS has been broken everywhere that whole time.
Is there anything that could have been done to stop ECS from working? I have inherited the account so there may be changes that I am not aware of
Might add that it works fine in EC2 mode, it is just Fargate that does not seem to work
Thanks Josh
This looks likely to be related to the service linked role AWSServiceRoleForECS
. If not already created, the creation during CreateCluster uses a best effort approach. In the result, service linked role AWSServiceRoleForECS
doesn't appear though ECS cluster is created.
For example, creating Service Linked role would fail due to throttling error for iam:CreateServiceLinkedRole.
I'd recommend you to create the Service Linked role AWSServiceRoleForECS
manually using CreateServiceLinkedRole API. You can also see the related documentation here.
Contenido relevante
- OFICIAL DE AWSActualizada hace un año
- OFICIAL DE AWSActualizada hace 4 meses
- OFICIAL DE AWSActualizada hace 3 años
- OFICIAL DE AWSActualizada hace 3 años
Well, after 3 days of it not working this suddenly sprang to life. I can only think that it WAS a problem on the AWS side and just took an inordinately long time to resolve. Thanks for all the pointers