Skip to main content
In this guide, we’ll explore how to work with AWS Lambda versions and aliases. You’ll learn how to create a new Lambda function, update its code to generate multiple versions, and use aliases to manage traffic between these versions effectively.

​
Creating the Initial Lambda Function (Version 1)

Begin by creating a new Lambda function. For this demonstration, the function is named “demo Lambda.” Once you access the AWS Lambda function creation page, you’ll see a setup screen similar to the one below:
The image shows the AWS Lambda function creation page, where a function named "demo-lambda" is being set up with Node.js 20.x runtime and x86_64 architecture.
Next, update the function code to indicate that this is the first version of your code:
Deploy the function and run a test. The test output should confirm that version one is active, displaying a response similar to:
When satisfied with version one, publish it by selecting “Publish new version” from the Actions menu. Use a description such as “This is v1” when prompted. AWS Lambda will then assign this as version 1 and redirect you to a page with the version details.
The image shows an AWS Lambda console with a function named "demo-lambda" at version 1. It includes options for adding triggers, destinations, and editing code.
When viewing a specific version, note that the function code is read-only. To modify the code, return to the main Lambda page.

​
Updating the Lambda Function (Creating Version 2 and Version 3)

Return to the main Lambda page to update the function code and create new versions.

​
Creating Version 2

For version two, update the code—even if it appears the same during testing—and deploy the changes:
After testing and verifying the output, publish these changes as a new version with a description like “version two.” This action creates version two, which remains immutable from that point onward.

​
Creating Version 3

Next, modify the code again to craft version three:
Deploy and publish this update as version three. Once complete, the main Lambda page will display all three available versions.

​
Creating and Managing Aliases

Aliases simplify the integration of Lambda functions with services such as API Gateway by abstracting specific version numbers.

​
1. Creating the “prod” Alias

Create an alias named “prod” to signify the production environment. Assign version one to the “prod” alias. The alias creation interface appears as follows:
The image shows an AWS Lambda interface for creating an alias, with fields for name, description, and version selection. The alias name is set to "prod."
Once saved, running a test should confirm the alias invokes version one.

​
2. Creating the “staging” Alias

Next, create an alias named “staging” and point it to version two. After creation, the Lambda console will show the alias:
The image shows an AWS Lambda console with an alias named "staging" pointing to a function called "demo-lambda." A green notification bar indicates the alias was successfully created and is pointing to version 2.
Testing verifies that the “staging” alias correctly invokes version two.

​
3. Creating the “dev” Alias

Similarly, set up a “dev” alias and assign it to version three. Tests should confirm that the “dev” alias invokes version three.

​
4. Creating a “playground” Alias for the Latest Code

If you haven’t published your latest changes, the main Lambda page shows the most recent code as “$LATEST.” You can create an alias—for example, “playground”—to point to this version. Before testing the “playground” alias, update the code to represent a new iteration:
Deploy these changes. Running tests on the “playground” alias should now display the new output. The AWS Lambda console showing multiple aliases looks like this:
The image shows an AWS Lambda console interface with a function named "demo-lambda." It displays details about the function, including aliases like "dev," "playground," "prod," and "staging."

​
Implementing Weighted Aliases

Weighted aliases enable you to gradually shift traffic between different function versions. For example, to roll out version two in production, you can allocate a specific percentage of traffic to it while maintaining the remaining traffic on version one. Follow these steps to set up weighted aliases:
  1. Navigate to the “prod” alias and click on “Edit.”
  2. Enable the weighted alias option. Add version two as an additional target and set the weight to 50% (in practice, you could start with a smaller percentage, such as 5-10%, for initial testing).
The traffic distribution interface appears as follows:
The image shows an AWS Lambda interface for editing an alias, where traffic is being distributed between two versions with specified weights. Version 1 has a weight of 95%, and an additional version 2 has a weight of 5%.
After saving, the “prod” alias will route 50% of the traffic to version one and the other 50% to version two. Repeated tests should yield alternating outputs; for instance:
This behavior indicates that half of the requests are handled by version two while the rest remain on version one.

​
Conclusion

In this guide, we’ve detailed how to create multiple versions of an AWS Lambda function and manage them using aliases. By leveraging aliases and weighted traffic distribution, you can integrate your Lambda functions seamlessly with services like API Gateway and roll out new versions without causing production disruptions. Happy coding!

Watch Video