Step 4: Deploy your Hello World application

First, the deployment needs a deployment group. However, before you create the deployment group, you need a service role ARN. A service role is an IAM role that gives a service permission to act on your behalf. In this case, the service role gives CodeDeploy permission to access your Amazon EC2 instances to expand (read) their Amazon EC2 instance tags.

You should have already followed the instructions in Create a service role (CLI) to create a service role. To get the ARN of the service role, see Get the service role ARN (CLI) .

Now that you have the ARN, call the create-deployment-group command to create a deployment group named HelloWorld_DepGroup, associated with the application named HelloWorld_App, using the Amazon EC2 instance tag named CodeDeployDemo and deployment configuration named CodeDeployDefault.OneAtATime, with the service role ARN:

aws deploy create-deployment-group --application-name HelloWorld_App --deployment-group-name HelloWorld_DepGroup --deployment-config-name CodeDeployDefault.OneAtATime --ec2-tag-filters Key=Name,Value=CodeDeployDemo,Type=KEY_AND_VALUE --service-role-arn serviceRoleARN

Before you create a deployment, the instances in your deployment group must have the CodeDeploy agent installed. You can install the agent from the command line with Amazon Systems Manager with the following command:

aws ssm create-association --name AWS-ConfigureAWSPackage --targets Key=tag:Name,Values=CodeDeployDemo --parameters action=Install,name=AWSCodeDeployAgent --schedule-expression "cron(0 2 ? * SUN *)" 

This command creates an association in Systems Manager State Manager that will install the CodeDeploy agent and then attempt to update it at 2:00 every Sunday morning. For more information about the CodeDeploy agent, see Working with the CodeDeploy agent. For more information about Systems Manager, see What is Amazon Systems Manager.

Now call the create-deployment command to create a deployment associated with the application named HelloWorld_App, the deployment configuration named CodeDeployDefault.OneAtATime, and the deployment group named HelloWorld_DepGroup, using the application revision named HelloWorld_App.zip in the bucket named codedeploydemobucket:

aws deploy create-deployment --application-name HelloWorld_App --deployment-config-name CodeDeployDefault.OneAtATime --deployment-group-name HelloWorld_DepGroup --s3-location bucket=codedeploydemobucket,bundleType=zip,key=HelloWorld_App.zip

Before you use the CodeDeploy console to deploy your application revision, you need a service role ARN. A service role is an IAM role that gives a service permission to act on your behalf. In this case, the service role gives CodeDeploy permission to access your Amazon EC2 instances to expand (read) their Amazon EC2 instance tags.

You should have already followed the instructions in Create a service role (console) to create a service role. To get the ARN of the service role, see Get the service role ARN (console) .

Now that you have the ARN, you can use the CodeDeploy console to deploy your application revision.

Sign in to the Amazon Web Services Management Console and open the CodeDeploy console at https://console.amazonaws.cn/codedeploy.

In the navigation pane, expand Deploy, then choose Applications.

Choose HelloWorld_App.

On the Deployment groups tab, choose Create deployment group.

In Deployment group name, enter HelloWorld_DepGroup.

In Service Role, choose the name of the service role.

In Deployment type, choose In-place.

In Environment configuration, select Amazon EC2 instances.

In Agent configuration with Amazon Systems Manager, keep the defaults.

In Key, enter Name.

In Value, enter CodeDeployDemo.

In Deployment configuration, choose CodeDeployDefault.OneAtATime.

In Load Balancer, clear Enable load balancing.

Choose Create deployment group.

Choose Create deployment.

In Deployment group, choose HelloWorld_DepGroup

In Revision type, choose My application is stored in Amazon S3, and then in Revision location, enter the location of the sample Hello World application revision you previously uploaded to Amazon S3. To get the location:

1. Open the Amazon S3 console at https://console.amazonaws.cn/s3/.

2. In the list of buckets, choose codedeploydemobucket (or the name of the bucket where you uploaded your application revision).

3. In the list of objects, choose HelloWorld_App.zip.

4. In the Overview tab, choose Copy path.

5. Return to the CodeDeploy console, and in Revision Location, paste the Link field value.

For Revision file type, choose .zip.

(Optional) Enter a comment in Deployment description.

Choose Create deployment. Information about your newly created deployment appears on the Deployments page.

Get the deployment's ID by calling the list-deployments command against the application named HelloWorld_App and the deployment group named HelloWorld_DepGroup:

aws deploy list-deployments --application-name HelloWorld_App --deployment-group-name HelloWorld_DepGroup --query "deployments" --output text

Call the get-deployment command with the deployment ID:

aws deploy get-deployment --deployment-id deploymentID --query "deploymentInfo.status" --output text

The command returns the deployment's overall status. If successful, the value is Succeeded.

If the overall status is Failed, you can call commands such as list-deployment-instances and get-deployment-instance to troubleshoot. For more troubleshooting options, see Analyzing log files to investigate deployment failures on instances.

In the Deployments table, choose your deployment ID. After a deployment fails, a message that describes the reason for the failure appears in the deployment's details page.

. More information about the deployment's instances is displayed. After a deployment fails, you might be able to determine on which Amazon EC2 instances and at which step the deployment failed.

If you want to do more troubleshooting, you can use a technique like View Instance Details. You can also analyze the deployment log files on an Amazon EC2 instance. For more information, see Analyzing log files to investigate deployment failures on instances.