Creating a Lambda function for a custom Amazon AppConfig extension - Amazon AppConfig
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Creating a Lambda function for a custom Amazon AppConfig extension

For most use-cases, to create a custom extension, you must create an Amazon Lambda function to perform any computation and processing defined in the extension. This section includes Lambda function sample code for a custom Amazon AppConfig extension. This section also includes payload request and response reference details. For information about creating a Lambda function, see Getting started with Lambda in the Amazon Lambda Developer Guide.

Sample code

The following sample code for a Lambda function, when invoked, automatically backs up an Amazon AppConfig configuration to an Amazon S3 bucket. The configuration is backed up whenever a new configuration is created or deployed. The sample uses extension parameters so the bucket name doesn't have to be hardcoded in the Lambda function. By using extension parameters, the user can attach the extension to multiple applications and back up configurations to different buckets. The code sample includes comments to further explain the function.

Sample Lambda function for an Amazon AppConfig extension

from datetime import datetime import base64 import json import boto3 def lambda_handler(event, context): print(event) # Extensions that use the PRE_CREATE_HOSTED_CONFIGURATION_VERSION and PRE_START_DEPLOYMENT # action points receive the contents of AWS AppConfig configurations in Lambda event parameters. # Configuration contents are received as a base64-encoded string, which the lambda needs to decode # in order to get the configuration data as bytes. For other action points, the content # of the configuration isn't present, so the code below will fail. config_data_bytes = base64.b64decode(event["Content"]) # You can specify parameters for extensions. The CreateExtension API action lets you define # which parameters an extension supports. You supply the values for those parameters when you # create an extension association by calling the CreateExtensionAssociation API action. # The following code uses a parameter called S3_BUCKET to obtain the value specified in the # extension association. You can specify this parameter when you create the extension # later in this walkthrough. extension_association_params = event.get('Parameters', {}) bucket_name = extension_association_params['S3_BUCKET'] write_backup_to_s3(bucket_name, config_data_bytes) # The PRE_CREATE_HOSTED_CONFIGURATION_VERSION and PRE_START_DEPLOYMENT action points can # modify the contents of a configuration. The following code makes a minor change # for the purposes of a demonstration. old_config_data_string = config_data_bytes.decode('utf-8') new_config_data_string = old_config_data_string.replace('hello', 'hello!') new_config_data_bytes = new_config_data_string.encode('utf-8') # The lambda initially received the configuration data as a base64-encoded string # and must return it in the same format. new_config_data_base64string = base64.b64encode(new_config_data_bytes).decode('ascii') return { 'statusCode': 200, # If you want to modify the contents of the configuration, you must include the new contents in the # Lambda response. If you don't want to modify the contents, you can omit the 'Content' field shown here. 'Content': new_config_data_base64string } def write_backup_to_s3(bucket_name, config_data_bytes): s3 = boto3.resource('s3') new_object = s3.Object(bucket_name, f"config_backup_{}.txt") new_object.put(Body=config_data_bytes)

If you want to use this sample during this walkthrough, save it with the name MyS3ConfigurationBackUpExtension and copy the Amazon Resource Name (ARN) for the function. You specify the ARN when you create the Amazon Identity and Access Management (IAM) assume role in the next section. You specify the ARN and the name when you create the extension.

Payload reference

This section includes payload request and response reference details for working with custom Amazon AppConfig extensions.

Request structure


{ 'InvocationId': 'vlns753', // id for specific invocation 'Parameters': { 'ParameterOne': 'ValueOne', 'ParameterTwo': 'ValueTwo' }, 'ContentType': 'text/plain', 'ContentVersion': '2', 'Content': 'SGVsbG8gZWFydGgh', // Base64 encoded content 'Application': { 'Id': 'abcd123', 'Name': 'ApplicationName' }, 'ConfigurationProfile': { 'Id': 'ijkl789', 'Name': 'ConfigurationName' }, 'Description': '', 'Type': 'PreCreateHostedConfigurationVersion', 'PreviousContent': { 'ContentType': 'text/plain', 'ContentVersion': '1', 'Content': 'SGVsbG8gd29ybGQh' } }


{ 'InvocationId': '765ahdm', 'Parameters': { 'ParameterOne': 'ValueOne', 'ParameterTwo': 'ValueTwo' }, 'ContentType': 'text/plain', 'ContentVersion': '2', 'Content': 'SGVsbG8gZWFydGgh', 'Application': { 'Id': 'abcd123', 'Name': 'ApplicationName' }, 'Environment': { 'Id': 'ibpnqlq', 'Name': 'EnvironmentName' }, 'ConfigurationProfile': { 'Id': 'ijkl789', 'Name': 'ConfigurationName' }, 'DeploymentNumber': 2, 'Description': 'Deployment description', 'Type': 'PreStartDeployment' }
Asynchronous events

OnStartDeployment, OnDeploymentStep, OnDeployment

{ 'InvocationId': 'o2xbtm7', 'Parameters': { 'ParameterOne': 'ValueOne', 'ParameterTwo': 'ValueTwo' }, 'Type': 'OnDeploymentStart', 'Application': { 'Id': 'abcd123' }, 'Environment': { 'Id': 'efgh456' }, 'ConfigurationProfile': { 'Id': 'ijkl789', 'Name': 'ConfigurationName' }, 'DeploymentNumber': 2, 'Description': 'Deployment description', 'ConfigurationVersion': '2' }
Response structure

The following examples show what your Lambda fuction returns in response to the request from a custom Amazon AppConfig extension.

Synchronous events - successful response

If you want to transform the content, use the following:

"Content": "SomeBase64EncodedByteArray"

If you don't want to transform the content, return nothing.

Asynchronous events - successful response

Return nothing.

All error events

{ "Error": "BadRequestError", "Message": "There was malformed stuff in here", "Details": [{ "Type": "Malformed", "Name": "S3 pointer", "Reason": "S3 bucket did not exist" }] }