DVA-C02 Deployment Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Update the function code using update-function-code.

  Why it's wrong here

  This method directly modifies the $LATEST version of the Lambda function. While it updates the code, it does not provide a mechanism for zero-downtime deployments or easy rollbacks. During the update, invocations might experience brief cold starts or inconsistent behavior as the underlying execution environment is refreshed, making it unsuitable for production environments requiring continuous availability.

• ✓

  Use a weighted alias to gradually shift traffic to a new version.

  Why this is correct

  This deployment method enables a canary release strategy, ensuring zero downtime. A new Lambda function version is published, and an alias is configured to distribute traffic between the existing stable version and the new version based on specified weights (e.g., 90% old, 10% new). This allows for gradual rollout, real-time monitoring of the new version's performance, and immediate rollback by adjusting weights if issues arise.

• ✓

  Create a new version and update the alias to point to the new version.

  Why this is correct

  This approach facilitates an atomic, blue/green deployment strategy. A new, immutable version of the Lambda function is published with the updated code. An existing alias, which previously pointed to the old stable version, is then atomically updated to point entirely to this new version. This instantly switches all traffic to the new version without any downtime, providing a clean cutover.

• ✗

  Create a new Lambda function and delete the old one.

  Why it's wrong here

  This method is highly disruptive and not recommended for updating an existing function without downtime. It involves creating an entirely new Lambda function resource, requiring all event sources (e.g., API Gateway, S3, SQS) and downstream consumers to be reconfigured to point to the new function's ARN. During this transition, the old function might still receive invocations while the new one is not yet fully configured, leading to significant downtime, invocation failures, and operational complexity.

• ✗

  Update the function configuration to increase memory.

  Why it's wrong here

  While increasing memory is a valid configuration change for a Lambda function, it does not constitute a method for updating the function's code. This action only modifies the allocated resources for the function's execution environment, impacting performance and cost, but it does not deploy new code or a new version of the application logic. The question specifically asks about methods to update the function itself, implying code changes.

Go deeper

Related to this question