Description
Invoking Lambda functions asynchronously from within the function code hides the application topology inside the code. An example might look as follows:
const aws = require('aws-sdk');
const lambda = new aws.Lambda();
exports.handler = async function(event, context) {
...
// Prepare invokation params
var params = {
FunctionName: process.env.FUNCTION_NAME,
InvocationType: Event,
Payload: JSON.stringify(event),
};
// Invoke the destination function
await lambda.invoke(params).promise();
...
};
Solution
Use Lambda Destination to send the message asynchronously and configure the Lambda destination in CloudFormation or CDK to make the application topology explicit.
See the CDK implementation of this pattern in Typescript on GitHub.
Considerations
Advantages
- Extracting the function invocation into a destination makes the application topology explicit and separates compsition from application logic
Applicability
Lambda Destinations are triggered only for asynchronous invocations. Therefore, this refactoring cannot be applied to synchronous invocations, e.g. from the console (Test). If your functions returns a synchronous result, this is now replaced by the message to be sent
Lambda Destinations are triggered at the end of the function execution whereas the code could send the message at any place
Invoking a functinon from AWS SDK allows you to set additional attributes like
ClientContext,InvocationType,LogType,Qualifier. You cannot set these attributes from a Lambda destination. SettingInvocationTypefor example allows for synchronous invocationsLambda Destinations have the option to send the response payload only via
responeOnly, not the full record. This allows to easily chain asynchronous Lambda functions without having to deal with data extraction in the runtime code. (LambdaDestinationOptions)Consider building event driven architectures using the other Destination types (SNS, SQS, EventBridge)
Related Refactorings
Related Resources
Created by:
