Deploy with AWS CodeDeploy from GitHub

Code deploy is a a fully managed deployment service which helps you deploy on AWS EC2, AWS ECS , AWS Fargate, AWS Lambda and on premises servers. This helps in faster releases with zero downtime.

Aim :

To deploy the code stored in a GitHub repository to multiple AWS instance at once.

Architecture

Sample Repository : https://github.com/kaumudigupta/GithubCodeDeployDemo

Requirements :

1. AWS EC2 instances with IAM role attached.
2. Access to GitHub repository where the code is present.
3. AWS IAM role for the AWS CodeDeploy Service.
4. Appspec file and scripts.

IAM Roles (IAM.txt):

a) IAM Role for Codedeploy (CodeDeployRole)

1. Policy Role for Code Deploy (AWSCodeDeployRole)

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Action": [
        "autoscaling:CompleteLifecycleAction",
  "autoscaling:DeleteLifecycleHook",
  "autoscaling:DescribeAutoScalingGroups",
  "autoscaling:DescribeLifecycleHooks",
  "autoscaling:PutLifecycleHook",
  "autoscaling:RecordLifecycleActionHeartbeat",
  "autoscaling:CreateAutoScalingGroup",
  "autoscaling:UpdateAutoScalingGroup",
  "autoscaling:EnableMetricsCollection",
  "autoscaling:DescribePolicies",
  "autoscaling:DescribeScheduledActions",
  "autoscaling:DescribeNotificationConfigurations",
  "autoscaling:SuspendProcesses",
  "autoscaling:ResumeProcesses",
  "autoscaling:AttachLoadBalancers",
  "autoscaling:AttachLoadBalancerTargetGroups",
  "autoscaling:PutScalingPolicy",
  "autoscaling:PutScheduledUpdateGroupAction",
  "autoscaling:PutNotificationConfiguration",
  "autoscaling:PutWarmPool",
  "autoscaling:DescribeScalingActivities",
  "autoscaling:DeleteAutoScalingGroup",
  "ec2:DescribeInstances",
  "ec2:DescribeInstanceStatus",
  "ec2:TerminateInstances",
  "tag:GetResources",
  "sns:Publish",
  "cloudwatch:DescribeAlarms",
  "cloudwatch:PutMetricAlarm",
  "elasticloadbalancing:DescribeLoadBalancers",
  "elasticloadbalancing:DescribeInstanceHealth",
  "elasticloadbalancing:RegisterInstancesWithLoadBalancer",
  "elasticloadbalancing:DeregisterInstancesFromLoadBalancer",
  "elasticloadbalancing:DescribeTargetGroups",
  "elasticloadbalancing:DescribeTargetHealth",
  "elasticloadbalancing:RegisterTargets",
  "elasticloadbalancing:DeregisterTargets"
      ],
      "Effect": "Allow",
      "Resource": "*"
    }
  ]
}

2. Trust policy for CodeDeploy :

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Sid": "",
      "Effect": "Allow",
      "Principal": {
        "Service": [
          "codedeploy.us-east-1.amazonaws.com"
        ]
      },
      "Action": "sts:AssumeRole"
    }
  ]
}

b) Instance Role for EC2 Instance (EC2DeployRole)

1. AWS Managed Role (AmazonEC2RoleforAWSCodeDeploy) :

{ 
    "Version": "2012-10-17", 
    "Statement": [   
      {     
          "Action": [       
              "s3:GetObject",
              "s3:GetObjectVersion",
              "s3:ListBucket"   
          ],     
          "Effect": "Allow",     
          "Resource": "*"   
      } 
    ]
}

Steps :

1. Upload the code on a private Github repository.
2. Create appspec.yml file to cater the needs of our code and upload it with the required scripts.
3. Create the IAM roles for the EC2 instance (EC2DeployRole) and for the CodeDeploy application. (CodeDeployDemo).
4. Create the AWS EC2 instances on which you want to deploy the application and use EC2DeployRole as IAM role during the launch.
5. Use a tag for the grouping of the instances.
6. Install and configure AWS CLI and then, install the CodeDeploy agent using SSH. Check codedeployagent.txt for steps in GitHub repo.
7. Use AWS CodeDeploy to create an application and a deployment group. Use in-place deployment and choose the instances to be put in the deployment group using the tag.
8. Choose “CodeDeployDemo” which we created earlier as the service role while creating and configuring the application.
9. We can choose if we want to deploy the app to one instance at a time or to half of the instances or to all of the instances in the deployment group.
10. Create new revision to deploy the application and provide the Github account credentials, the name of the repository and the commit ID.
11. It will start deploying the application and to deploy a new commit in the future, just create a new revision.

Conclusion :

We can deploy our code easily using CodeDeploy on ’n’ number of instances.