aws_ecs_task_definition ClientException: Too many concurrent attempts to create a new revision of the specified family. #258
Description
Describe the Feature
When destroy and create task definition, about 1 in 10 we get the below error. Adding skip_destroy variable to aws_ecs_task_definition resource is a workaround to this issue.
Terraform will perform the following actions:
# aws_ecs_task_definition.service must be replaced
# module.ecs_web.module.ecs_alb_service_task.aws_ecs_service.default[0] will be updated in-place
# module.ecs_web.module.ecs_alb_service_task.aws_ecs_task_definition.default[0] must be replaced
Plan: 2 to add, 1 to change, 2 to destroy.
Changes to Outputs:
module.ecs_web.module.ecs_alb_service_task.aws_ecs_task_definition.default[0]: Destroying... [id=team-infra-worker]
aws_ecs_task_definition.service: Destroying... [id=team-infra-worker]
aws_ecs_task_definition.service: Destruction complete after 1s
module.ecs_web.module.ecs_alb_service_task.aws_ecs_task_definition.default[0]: Destruction complete after 1s
aws_ecs_task_definition.service: Creating...
module.ecs_web.module.ecs_alb_service_task.aws_ecs_task_definition.default[0]: Creating...
aws_ecs_task_definition.service: Creation complete after 0s [id=team-infra-worker]
│ Warning: Argument is deprecated
│
│ with module.ecs_web.module.ecs_codepipeline.aws_s3_bucket.default,
│ on .terraform/modules/ecs_web.ecs_codepipeline/main.tf line 16, in resource "aws_s3_bucket" "default":
│ 16: acl = "private"
│
│ Use the aws_s3_bucket_acl resource instead
│ Error: creating ECS Task Definition (team-infra-worker): operation error ECS: RegisterTaskDefinition, https response error StatusCode: 400, RequestID: XXXXXXX, ClientException: Too many concurrent attempts to create a new revision of the specified family.
│
│ with module.ecs_web.module.ecs_alb_service_task.aws_ecs_task_definition.default[0],
│ on .terraform/modules/ecs_web.ecs_alb_service_task/main.tf line 39, in resource "aws_ecs_task_definition" "default":
│ 39: resource "aws_ecs_task_definition" "default" {
│
Error: Process completed with exit code 1.
Expected Behavior
destroy and create of new task definition should complete without errors
Use Case
Prevent random terraform error
Describe Ideal Solution
Ideal solution is for terraform to handle a destroy and immediate creation of the task definition reliably without errors
Alternatives Considered
Adding skip_destroy variable to aws_ecs_task_definition resource is a workaround to this issue as the destroy will never happen
Additional Context
No response