# Tutorial: Create a pipeline with Amazon CloudFormation
The examples provide sample templates that allow you to use Amazon CloudFormation to create a pipeline that deploys your application to your instances each time the source code changes. The sample template creates a pipeline that you can view in Amazon CodePipeline. The pipeline detects the arrival of a saved change through Amazon CloudWatch Events.
**Important**
As part of creating a pipeline, an S3 artifact bucket provided by the customer will be used by CodePipeline for artifacts. (This is different from the bucket used for an S3 source action.) If the S3 artifact bucket is in a different account from the account for your pipeline, make sure that the S3 artifact bucket is owned by Amazon Web Services accounts that are safe and will be dependable.
The CodePipeline action for Amazon CloudFormation is not available in the China (Beijing) and China (Ningxia) Regions. To reference other available actions, see [Product and service integrations with CodePipeline](integrations.md).
**Topics**
+ [
# Example 1: Create an Amazon CodeCommit pipeline with Amazon CloudFormation
](tutorials-cloudformation-codecommit.md)
+ [
# Example 2: Create an Amazon S3 pipeline with Amazon CloudFormation
](tutorials-cloudformation-s3.md)
# Example 1: Create an Amazon CodeCommit pipeline with Amazon CloudFormation
This walkthrough shows you how to use the Amazon CloudFormation console to create infrastructure that includes a pipeline connected to a CodeCommit source repository. In this tutorial, you use the provided sample template file to create your resource stack, which includes your artifact store, pipeline, and change-detection resources, such as your Amazon CloudWatch Events rule. After you create your resource stack in Amazon CloudFormation, you can view your pipeline in the Amazon CodePipeline console. The pipeline is a two-stage pipeline with a CodeCommit source stage and a CodeDeploy deployment stage.
**Prerequisites:**
You must have created the following resources to use with the Amazon CloudFormation sample template:
+ You must have created a source repository. You can use the Amazon CodeCommit repository you created in [Tutorial: Create a simple pipeline (CodeCommit repository)](tutorials-simple-codecommit.md).
+ You must have created a CodeDeploy application and deployment group. You can use the CodeDeploy resources you created in [Tutorial: Create a simple pipeline (CodeCommit repository)](tutorials-simple-codecommit.md).
+ Choose one of these links to download the sample Amazon CloudFormation template file for creating a pipeline: [YAML](samples/codepipeline-codecommit-events-yaml.zip) \$1 [JSON](samples/codepipeline-codecommit-events-json.zip)
Unzip the file and place it on your local computer.
+ Download the [SampleApp\$1Linux.zip](samples/SampleApp_Linux.zip) sample application file.
**Create your pipeline in Amazon CloudFormation**
1. Unzip the files from [SampleApp\$1Linux.zip](samples/SampleApp_Linux.zip) and upload the files to your Amazon CodeCommit repository. You must upload the unzipped files to the root directory of your repository. You can follow the instructions in [Step 2: Add sample code to your CodeCommit repository](tutorials-simple-codecommit.md#codecommit-add-code) to push the files to your repository.
1. Open the Amazon CloudFormation console and choose **Create Stack**. Choose **With new resources (standard)**.
1. Under **Specify template**, choose **Upload a template**. Select **Choose file** and then choose the template file from your local computer. Choose **Next**.
1. In **Stack name**, enter a name for your pipeline. Parameters specified by the sample template are displayed. Enter the following parameters:
1. In **ApplicationName**, enter the name of your CodeDeploy application.
1. In **BetaFleet**, enter the name of your CodeDeploy deployment group.
1. In **BranchName**, enter the repository branch you want to use.
1. In **RepositoryName**, enter the name of your CodeCommit source repository.
1. Choose **Next**. Accept the defaults on the following page, and then choose **Next**.
1. In **Capabilities**, select **I acknowledge that Amazon CloudFormation might create IAM resources**, and then choose **Create stack**.
1. After your stack creation is complete, view the event list to check for any errors.
**Troubleshooting**
The IAM user who is creating the pipeline in Amazon CloudFormation might require additional permissions to create resources for the pipeline. The following permissions are required in the policy to allow Amazon CloudFormation to create the required Amazon CloudWatch Events resources for the CodeCommit pipeline:
```
{
"Effect": "Allow",
"Action": [
"events:PutRule",
"events:PutEvents",
"events:PutTargets",
"events:DeleteRule",
"events:RemoveTargets",
"events:DescribeRule"
],
"Resource": "resource_ARN"
}
```
1. Sign in to the Amazon Web Services Management Console and open the CodePipeline console at [https://console.amazonaws.cn/codepipeline/](https://console.amazonaws.cn/codepipeline/).
Under **Pipelines**, choose your pipeline and choose **View**. The diagram shows your pipeline source and deployment stages.
**Note**
To view the pipeline that was created, find the **Logical ID** column under the **Resources** tab for your stack in Amazon CloudFormation. Note the name in the **Physical ID** column for the pipeline. In CodePipeline, you can view the pipeline with the same Physical ID (pipeline name) in the Region where you created your stack.
1. In your source repository, commit and push a change. Your change-detection resources pick up the change, and your pipeline starts.
# Example 2: Create an Amazon S3 pipeline with Amazon CloudFormation
This walkthrough shows you how to use the Amazon CloudFormation console to create infrastructure that includes a pipeline connected to an Amazon S3 source bucket. In this tutorial, you use the provided sample template file to create your resource stack, which includes your source bucket, artifact store, pipeline, and change-detection resources, such as your Amazon CloudWatch Events rule and CloudTrail trail. After you create your resource stack in Amazon CloudFormation, you can view your pipeline in the Amazon CodePipeline console. The pipeline is a two-stage pipeline with an Amazon S3 source stage and a CodeDeploy deployment stage.
**Prerequisites:**
You must have the following resources to use with the Amazon CloudFormation sample template:
+ You must have created the Amazon EC2 instances, where you installed the CodeDeploy agent on the instances. You must have created a CodeDeploy application and deployment group. Use the Amazon EC2 and CodeDeploy resources you created in [Tutorial: Create a simple pipeline (CodeCommit repository)](tutorials-simple-codecommit.md).
+ Choose the following links to download the sample Amazon CloudFormation template files for creating a pipeline with an Amazon S3 source:
+ Download the sample template for your pipeline: [YAML](samples/codepipeline-s3-events-yaml.zip) \$1 [JSON](samples/codepipeline-s3-events-json.zip)
+ Download the sample template for your CloudTrail bucket and trail: [YAML](samples/codepipeline-s3-cloudtrail-yaml.zip) \$1 [JSON](samples/codepipeline-s3-cloudtrail-json.zip)
+ Unzip the files and place them on your local computer.
+ Download the sample application from [SampleApp\$1Linux.zip](samples/SampleApp_Linux.zip).
Save the .zip file on your local computer. You upload the .zip file after the stack is created.
**Create your pipeline in Amazon CloudFormation**
1. Open the Amazon CloudFormation console, and choose **Create Stack**. Choose **With new resources (standard)**.
1. In **Choose a template**, choose **Upload a template**. Select **Choose file**, and then choose the template file from your local computer. Choose **Next**.
1. In **Stack name**, enter a name for your pipeline. Parameters specified by the sample template are displayed. Enter the following parameters:
1. In **ApplicationName**, enter the name of your CodeDeploy application. You can replace the `DemoApplication` default name.
1. In **BetaFleet**, enter the name of your CodeDeploy deployment group. You can replace the `DemoFleet` default name.
1. In **SourceObjectKey**, enter `SampleApp_Linux.zip`. You upload this file to your bucket after the template creates the bucket and pipeline.
1. Choose **Next**. Accept the defaults on the following page, and then choose **Next**.
1. In **Capabilities**, select **I acknowledge that Amazon CloudFormation might create IAM resources**, and then choose **Create stack**.
1. After your stack creation is complete, view the event list to check for any errors.
**Troubleshooting**
The IAM userwho is creating the pipeline in Amazon CloudFormation might require additional permissions to create resources for the pipeline. The following permissions are required in the policy to allow Amazon CloudFormation to create the required Amazon CloudWatch Events resources for the Amazon S3 pipeline:
```
{
"Effect": "Allow",
"Action": [
"events:PutRule",
"events:PutEvents",
"events:PutTargets",
"events:DeleteRule",
"events:RemoveTargets",
"events:DescribeRule"
],
"Resource": "resource_ARN"
}
```
1. In Amazon CloudFormation, in the **Resources** tab for your stack, view the resources that were created for your stack.
**Note**
To view the pipeline that was created, find the **Logical ID** column under the **Resources** tab for your stack in Amazon CloudFormation. Note the name in the **Physical ID** column for the pipeline. In CodePipeline, you can view the pipeline with the same Physical ID (pipeline name) in the Region where you created your stack.
Choose the S3 bucket with a `sourcebucket` label in the name, such as `s3-cfn-codepipeline-sourcebucket-y04EXAMPLE.` Do not choose the pipeline artifact bucket.
The source bucket is empty because the resource is newly created by Amazon CloudFormation. Open the Amazon S3 console and locate your `sourcebucket` bucket. Choose **Upload**, and follow the instructions to upload your `SampleApp_Linux.zip` .zip file.
**Note**
When Amazon S3 is the source provider for your pipeline, you must upload to your bucket all source files packaged as a single .zip file. Otherwise, the source action fails.
1. Sign in to the Amazon Web Services Management Console and open the CodePipeline console at [https://console.amazonaws.cn/codepipeline/](https://console.amazonaws.cn/codepipeline/).
Under **Pipelines**, choose your pipeline, and then choose **View**. The diagram shows your pipeline source and deployment stages.
1. Complete the steps in the following procedure to create your Amazon CloudTrail resources.
**Create your Amazon CloudTrail resources in Amazon CloudFormation**
1. Open the Amazon CloudFormation console, and choose **Create Stack**.
1. In **Choose a template**, choose **Upload a template to Amazon S3**. Choose **Browse**, and then select the template file for the Amazon CloudTrail resources from your local computer. Choose **Next**.
1. In **Stack name**, enter a name for your resource stack. Parameters specified by the sample template are displayed. Enter the following parameters:
1. In **SourceObjectKey**, accept the default for the sample application's zip file.
1. Choose **Next**. Accept the defaults on the following page, and then choose **Next**.
1. In **Capabilities**, select **I acknowledge that Amazon CloudFormation might create IAM resources**, and then choose **Create**.
1. After your stack creation is complete, view the event list to check for any errors.
The following permissions are required in the policy to allow Amazon CloudFormation to create the required CloudTrail resources for the Amazon S3 pipeline:
```
{
"Effect": "Allow",
"Action": [
"cloudtrail:CreateTrail",
"cloudtrail:DeleteTrail",
"cloudtrail:StartLogging",
"cloudtrail:StopLogging",
"cloudtrail:PutEventSelectors"
],
"Resource": "resource_ARN"
}
```
1. Sign in to the Amazon Web Services Management Console and open the CodePipeline console at [https://console.amazonaws.cn/codepipeline/](https://console.amazonaws.cn/codepipeline/).
Under **Pipelines**, choose your pipeline, and then choose **View**. The diagram shows your pipeline source and deployment stages.
1. In your source bucket, commit and push a change. Your change-detection resources pick up the change and your pipeline starts.