Configuring the Dockerrun.aws.json v2 file

Dockerrun.aws.json v2 is an Elastic Beanstalk configuration file that describes how to deploy a set of Docker containers hosted in an ECS cluster in an Elastic Beanstalk environment. The Elastic Beanstalk platform creates an ECS task definition, which includes an ECS container definition. These definitions are described in the Dockerrun.aws.json configuration file.

The container definition in the Dockerrun.aws.json file describes the containers to deploy to each Amazon EC2 instance in the ECS cluster. In this case an Amazon EC2 instance is also referred to as a host container instance, because it hosts the Docker containers. The configuration file also describes the data volumes to create on the host container instance for the Docker containers to mount. For more information and a diagram of the components in an ECS managed Docker environment on Elastic Beanstalk, see the ECS managed Docker platform overview earlier in this chapter.

A Dockerrun.aws.json file can be used on its own or zipped up with additional source code in a single archive. Source code that is archived with a Dockerrun.aws.json is deployed to Amazon EC2 container instances and accessible in the /var/app/current/ directory.

Dockerrun.aws.json v2

The Dockerrun.aws.json file includes the following sections:

AWSEBDockerrunVersion

Specifies the version number as the value 2 for ECS managed Docker environments.

executionRoleArn

Specifies the task execution IAM roles for different purposes and services associated with your account. For your application to use Elastic Beanstalk environment variables stored as secrets, you’ll need to specify the ARN of a task execution role that grants the required permissions. Other common use cases may also require this parameter. For more information, see Execution Role ARN format.

volumes

Creates volumes from folders in the Amazon EC2 container instance, or from your source bundle (deployed to /var/app/current). Mount these volumes to paths within your Docker containers using mountPoints in the containerDefinitions section.

containerDefinitions

An array of container definitions.

authentication (optional)

The location in Amazon S3 of a .dockercfg file that contains authentication data for a private repository.

The containerDefinitions and volumes sections of Dockerrun.aws.json use the same formatting as the corresponding sections of an Amazon ECS task definition file. For more information about the task definition format and a full list of task definition parameters, see Amazon ECS task definitions in the Amazon Elastic Container Service Developer Guide.

Volume format

The volume parameter creates volumes from either folders in the Amazon EC2 container instance, or from your source bundle (deployed to /var/app/current).

Volumes are specified in the following format:

"volumes": [
    {
      "name": "volumename",
      "host": {
        "sourcePath": "/path/on/host/instance"
      }
    }
  ],

Mount these volumes to paths within your Docker containers using mountPoints in the container definition.

Elastic Beanstalk configures additional volumes for logs, one for each container. These should be mounted by your Docker containers in order to write logs to the host instance.

For more details, see the mountPoints field in the Container definition format section that follows.

Execution Role ARN format

In order for your application to use Elastic Beanstalk environment variables stored as secrets, you'll need to specify a task execution IAM role. The role must grant the Amazon ECS container permission to make AWS API calls on your behalf using AWS Secrets Manager secrets or AWS Systems Manager Parameter Store parameters to reference sensitive data. For instructions to create a task execution IAM role with the required permissions for your account, see Amazon ECS task execution IAM role in the Amazon Elastic Container Service Developer Guide.

{
"AWSEBDockerrunVersion": 2,
  "executionRoleArn": "arn:aws:iam::111122223333:role/ecsTaskExecutionRole",

Additional permissions required for the Amazon ECS managed Docker platform

EC2 instance profile grants iam:PassRole to ECS

In order for your EC2 instance profile to be able to grant this role to the ECS container, you must include the iam:PassRole permission demonstrated in the following example. The iam:PassRole allows the EC2 instances permission to pass the task execution role to the ECS container.

In this example, we limit the EC2 instance to only pass the role to the ECS service. Although this condition is not required, we add it to follow best practices to reduce the scope of the permission shared. We accomplish this with the Condition element.

Note

Any usage of the ECS IAM task execution role requires the iam:PassRole permission. There are other common use cases that require the ECS task execution managed service role. For more information, see Amazon ECS task execution IAM role in the Amazon Elastic Container Service Developer Guide.

Example policy with iam:PassRole permission

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": "iam:PassRole",
            "Resource": [
                "arn:aws:iam::123456789012:role/ecs-task-execution-role"
            ],
            "Condition": {
                "StringLike": {
                    "iam:PassedToService": "ecs-tasks.amazonaws.com"
                }
            }
        }
    ]
}

Granting secrets and parameters access to the Amazon ECS container agent

The Amazon ECS task execution IAM role also needs permissions to access the secrets and parameter stores. Similar to the requirements of the EC2 instance profile role, the ECS container agent requires permission to pull the necessary Secrets Manager or Systems Manager resources. For more information, see Secrets Manager or Systems Manager permissions in the Amazon Elastic Container Service Developer Guide

Granting secrets and parameters access to the Elastic Beanstalk EC2 instances

To support secrets configured as environment variables, you'll also need to add permissions to your EC2 instance profile. For more information, see Fetching secrets and parameters to Elastic Beanstalk environment variables and Required IAM permissions for Secrets Manager.

The following examples combine the previous iam:PassRole example with the examples provided in the referenced Required IAM permissions for Secrets Manager. They add the permissions that the EC2 instances require to access the AWS Secrets Manager and AWS Systems Manager stores to retrieve the secrets and parameter data to initialize the Elastic Beanstalk environment variables that have been configured for secrets.

Example Secrets Manager policy combined with iam:PassRole permission

{
    "Version": "2012-10-17",
    "Statement": [
       {
            "Effect": "Allow",
            "Action": "iam:PassRole",
            "Resource": [
                "arn:aws:iam::123456789012:role/ecs-task-execution-role"
            ],
            "Condition": {
                "StringLike": {
                    "iam:PassedToService": "ecs-tasks.amazonaws.com"
               }
            } 
        },
        {
            "Effect": "Allow",
            "Action": [
                "secretsmanager:GetSecretValue",
                "kms:Decrypt"          
            ],
            "Resource": [
                "arn:aws:secretsmanager:us-east-1:111122223333:secret:my-secret",
                "arn:aws:kms:us-east-1:111122223333:key/my-key"
            ]
        }
    ]
}

Example Systems Manager policy combined with iam:PassRole permission

{
    "Version": "2012-10-17",
    "Statement": [
       {
            "Effect": "Allow",
            "Action": "iam:PassRole",
            "Resource": [
                "arn:aws:iam::123456789012:role/ecs-task-execution-role"
            ],
            "Condition": {
                "StringLike": {
                    "iam:PassedToService": "ecs-tasks.amazonaws.com"
               }
            } 
        },
        {
            "Effect": "Allow",
            "Action": [
                "ssm:GetParameter",
                "kms:Decrypt"
            ],
            "Resource": [
                "arn:aws:ssm:us-east-1:111122223333:parameter/my-parameter",
                "arn:aws:kms:us-east-1:111122223333:key/my-key"
            ]
        }
    ]
}

Container definition format

The following examples show a subset of parameters that are commonly used in the containerDefinitions section. More optional parameters are available.

The Beanstalk platform creates an ECS task definition, which includes an ECS container definition. Beanstalk supports a sub-set of parameters for the ECS container definition. For more information, see Container definitions in the Amazon Elastic Container Service Developer Guide.

A Dockerrun.aws.json file contains an array of one or more container definition objects with the following fields:

name

The name of the container. See Standard Container Definition Parameters for information about the maximum length and allowed characters.

image

The name of a Docker image in an online Docker repository from which you're building a Docker container. Note these conventions:

• Images in official repositories on Docker Hub use a single name (for example, ubuntu or mongo).

• Images in other repositories on Docker Hub are qualified with an organization name (for example, amazon/amazon-ecs-agent.

• Images in other online repositories are qualified further by a domain name (for example, quay.io/assemblyline/ubuntu).

environment

An array of environment variables to pass to the container.

For example, the following entry defines an environment variable with the name Container and the value PHP:

"environment": [
  {
    "name": "Container",
    "value": "PHP"
  }
],

essential

True if the task should stop if the container fails. Nonessential containers can finish or crash without affecting the rest of the containers on the instance.

memory

Amount of memory on the container instance to reserve for the container. Specify a non-zero integer for one or both of the memory or memoryReservation parameters in container definitions.

memoryReservation

The soft limit (in MiB) of memory to reserve for the container. Specify a non-zero integer for one or both of the memory or memoryReservation parameters in container definitions.

mountPoints

Volumes from the Amazon EC2 container instance to mount, and the location on the Docker container file system at which to mount them. When you mount volumes that contain application content, your container can read the data you upload in your source bundle. When you mount log volumes for writing log data, Elastic Beanstalk can gather log data from these volumes.

Elastic Beanstalk creates log volumes on the container instance, one for each Docker container, at /var/log/containers/containername. These volumes are named awseb-logs-containername and should be mounted to the location within the container file structure where logs are written.

For example, the following mount point maps the nginx log location in the container to the Elastic Beanstalk–generated volume for the nginx-proxy container.

{
  "sourceVolume": "awseb-logs-nginx-proxy",
  "containerPath": "/var/log/nginx"
}

portMappings

Maps network ports on the container to ports on the host.

links

List of containers to link to. Linked containers can discover each other and communicate securely.

volumesFrom

Mount all of the volumes from a different container. For example, to mount volumes from a container named web:

"volumesFrom": [
  {
    "sourceContainer": "web"
  }
],

Authentication format – using images from a private repository

The authentication section contains authentication data for a private repository. This entry is optional.

Add the information about the Amazon S3 bucket that contains the authentication file in the authentication parameter of the Dockerrun.aws.json file. Make sure that the authentication parameter contains a valid Amazon S3 bucket and key. The Amazon S3 bucket must be hosted in the same region as the environment that is using it. Elastic Beanstalk will not download files from Amazon S3 buckets hosted in other regions.

Uses the following format:

"authentication": {
    "bucket": "amzn-s3-demo-bucket",
    "key": "mydockercfg"
  },

For information about generating and uploading the authentication file, see Using images from a private repository in Elastic Beanstalk.

Example Dockerrun.aws.json v2

The following snippet is an example that illustrates the syntax of the Dockerrun.aws.json file for an instance with two containers.

{
  "AWSEBDockerrunVersion": 2,
  "volumes": [
    {
      "name": "php-app",
      "host": {
        "sourcePath": "/var/app/current/php-app"
      }
    },
    {
      "name": "nginx-proxy-conf",
      "host": {
        "sourcePath": "/var/app/current/proxy/conf.d"
      }
    }
  ],
  "containerDefinitions": [
    {
      "name": "php-app",
      "image": "php:fpm",
      "environment": [
        {
          "name": "Container",
          "value": "PHP"
        }
      ],
      "essential": true,
      "memory": 128,
      "mountPoints": [
        {
          "sourceVolume": "php-app",
          "containerPath": "/var/www/html",
          "readOnly": true
        }
      ]
    },
    {
      "name": "nginx-proxy",
      "image": "nginx",
      "essential": true,
      "memory": 128,
      "portMappings": [
        {
          "hostPort": 80,
          "containerPort": 80
        }
      ],
      "links": [
        "php-app"
      ],
      "mountPoints": [
        {
          "sourceVolume": "php-app",
          "containerPath": "/var/www/html",
          "readOnly": true
        },
        {
          "sourceVolume": "nginx-proxy-conf",
          "containerPath": "/etc/nginx/conf.d",
          "readOnly": true
        },
        {
          "sourceVolume": "awseb-logs-nginx-proxy",
          "containerPath": "/var/log/nginx"
        }
      ]
    }
  ]
}