Drone on ECS

Author: Darren Coxall

This repository serves as an example of how to get a fully functioning Drone CI container running on AWS EC2 Container Service. The method selected for this is to use CloudFormation which makes it easy to cleanup your account afterwards, alter your setup and completely re-produce your environment when necessary.

What does the template do

The provided template will provision systems within your AWS account and these will have associated costs. Please check the template before running it so you understand what is being set-up and to avoid any surprises.

The template does the following:

• Defines a task for Drone forwarding host port 80 to container port 8000
• Creates an auto scaling group that ensures a single instance is running
• Creates a load balancer sends traffic on port 80 to the host port 80
• Defines an ECS cluster and service that links the load balancer, instance and task
• Applies security groups to the instance and load balancer to ensure we can:
  • SSH onto the instance if we need to debug
  • Prevents direct traffic (all HTTP must come from the load balancer)
• Creates necessary IAM roles to ensure instances can register to the cluster

Usage

Clone this repository or copy the template into a directory. Using the aws command line tools you can then follow the steps detailed below.

Create a file that will store our template parameters. These can be provided directly using the CLI but with a handful to define it gets a bit messy.

# params.json
[
    {
        "ParameterKey": "VPC",
        "ParameterValue": "[VPC ID]"
    },
    {
        "ParameterKey": "Subnets",
        "ParameterValue": "[CSV LIST OF SUBNET IDS]"
    },
    {
        "ParameterKey": "KeyName",
        "ParameterValue": "[KEYPAIR NAME]"
    },
    {
        "ParameterKey": "DroneRemoteDriver",
        "ParameterValue": "[DRONE REMOTE_DRIVER VALUE]"
    },
    {
        "ParameterKey": "DroneRemoteConfig",
        "ParameterValue": "[DRONE REMOTE_CONFIG VALUE]"
    }
]

There are also 4 optional parameters that can be overridden.

• DroneMemoryAllocation - How much memory do you want to dedicate to the Drone container
• DroneCpuUnits - How many CPU units will be assigned to the Drone container. See here for more information
• DroneInstanceType - The AWS instance type to use. Defaults to m4.large
• IncomingHttpCidr - A CIDR range to limit incoming HTTP traffic. Be careful changing this as GitHub needs to be able to communicate with the load balancer

For the most part, these can be left as is unless you wish to run smaller instances in which you can adjust the instance type and the memory and cpu allocation accordingly.

With your parameters defined you can now execute...

$ aws cloudformation create-stack --stack-name drone-ci \
        --template-body file://template.json \
        --parameters file://params.json \
        --capabilities CAPABILITY_IAM

In the AWS console you will now see cloudformation provisioning your new Drone setup. Once complete you can check the outputs of the stack to help you configure your remote driver (i.e. GitHub) with the correct oauth endpoint.

$ aws cloudformation describe-stacks --stack-name drone-ci \
        --query 'Stacks[0].Outputs[0].OutputValue' \
        --output text

This value is also the URL you can use to access your installation.

Contributing

See room for improvement? Have a suggesstion? or even discovered an issue or bug? Please feel welcome to contribute to the project with a pull request or an issue.

License

This project is licensed under an Apache License 2.0.