0.3.4

Fix Security Group ingress rules

what

• Removed the ingress rule from EC2 Security Group

ingress {
    from_port   = 0
    to_port     = 0
    protocol    = -1
    cidr_blocks = ["0.0.0.0/0"]
  }

why

• Bad security practice
• Too broad access from any IP address by default
• Access should be controlled by provided Security Groups instead of IP ranges
• Was overriding the next ingress rule for the same Security Group

ingress {
    from_port       = 0
    to_port         = 0
    protocol        = -1
    security_groups = ["${var.security_groups}"]
  }

0.3.3

what

• Bump terraform-null-label version to 0.3.1

why

• Latest version with bug fixes

0.3.2

what

• Added Travis CI
• terraform fmt

why

• To monitor repo status
• For terraform lint

Add variable for environment tier

0.3.1

Add variable for environment tier (#22)

0.3.0: Added timeout variable (#20)

What

• Added configurable timeout

Why

• For huge and complex envs 20 minutes to became ready is not enough

Allow elb proxy ssh

Add SSH Forwarding (#16)

* Added ssh listener

* Fix type cast

* Fix type cast

* Fix type cast

* Fix type cast

* Fix type cast

* Added option to manage ssh port

Add `ec2_instance_profile_role_name` to outputs

what

• Added ec2_instance_profile_role_name to outputs

why

• To be able to attach additional aws_iam_role_policy_attachment's to the instance profile role from other modules

0.2.7: Add notifications (#13)

* Adde notifications

* Added support of different notificaton targets

Add `AmazonEC2ContainerRegistryReadOnly` policy to provide read-only access to `ECR` repositories

What

• Add AmazonEC2ContainerRegistryReadOnly managed policy to EB

Why

• To provide read-only access to all Amazon ECR repositories in the account

• When CodePipeline builds and pushes Docker images to ECR and then deploys the file Dockerrun.aws.json to EB with the ECR repo URL specified, EB needs permissions to pull the Docker image from the ECR repo to deploy it to EC2 instances

References

• http://docs.aws.amazon.com/elasticbeanstalk/latest/dg/create_deploy_docker.container.console.html
• http://docs.aws.amazon.com/AmazonECR/latest/userguide/ecr_managed_policies.html#AmazonEC2ContainerRegistryReadOnly

Change custom ENV vars key/value calculation

What

• Changed the way the custom ENV vars are calculated in aws:elasticbeanstalk:application:environment setting

Why

• Using null_resource to generate key/value pairs for ENV vars like this:

resource "null_resource" "env_vars" {
  count = 50

  triggers {
    key = "${count.index < length(var.env_vars) ?
                        element(concat(keys(var.env_vars),list("")), count.index) :
                        format(var.env_default_key, count.index+1)
               }"

    value = "${count.index < length(var.env_vars) ?
                        lookup(var.env_vars, element(concat(keys(var.env_vars),list("")), count.index), var.env_default_value) :
                        var.env_default_value
                 }"
  }
}

setting {
  namespace = "aws:elasticbeanstalk:application:environment"
  name      = "${null_resource.env_vars.0.triggers.key}"
  value     = "${null_resource.env_vars.0.triggers.value}"
}
setting {
  namespace = "aws:elasticbeanstalk:application:environment"
  name      = "${null_resource.env_vars.1.triggers.key}"
  value     = "${null_resource.env_vars.1.triggers.value}"
}

DOES NOT work when it's used from top-level modules AND when there is a dependency between modules AND terraform apply or terraform plan runs for the first time (meaning no AWS resources have been created yet), e.g.

module "elastic_beanstalk_environment" {
  source        = "git::https://github.com/cloudposse/terraform-aws-elastic-beanstalk-environment.git?ref=tags/0.2.4"

  env_vars                = "${
      merge(
        map(
          "EFS_HOST", "${module.efs.dns_name}"
        ), var.env_vars
      )
    }"
}

module "efs" {
  source             = "git::https://github.com/cloudposse/terraform-aws-efs.git?ref=tags/0.3.1"
  namespace          = "${var.namespace}"
  name               = "${var.name}"
  stage              = "${var.stage}"
  security_groups    = ["${module.elastic_beanstalk_environment.security_group_id}"]
}

Terraform throws the error:

the key for null_resource.env_vars.0.triggers.key could not be found

Looks like Terraform can't resolve the module inter-dependencies when using null_resource.
The proposed way of generating key/value pairs for ENV vars works in all cases.