DOP-C02 Question 89
Select 3Your team is developing a CI/CD pipeline for a serverless application using AWS Lambda. The pipeline must ensure that new code deployments to the Lambda function do not degrade application performance. Which approach should you implement in the pipeline to achieve this requirement?
- A
Use AWS CodeDeploy to configure a Canary deployment strategy for the Lambda function.
- B
Integrate Amazon CloudWatch alarms to roll back changes if a performance metric exceeds a threshold.
- C
Manually test the Lambda function after deployment to ensure it performs as expected.
- D
Use CodePipeline to implement a pre-production testing stage with automated performance tests.
- E
Deploy the new version of the Lambda function directly to production to verify its performance under real user traffic.
Show answer and explanation
Correct answers: A, B, D
Explanation
To ensure that new code deployments to a Lambda function do not degrade application performance, it is essential to employ best practices such as Canary deployments, automated performance testing, and monitoring with rollback mechanisms. AWS CodeDeploy's Canary strategy and CloudWatch alarms provide mechanisms to gradually roll out changes and detect issues early. Automated testing in a pre-production stage adds another layer of validation to ensure performance standards are met prior to production deployment.
- A. Correct.
Correct: AWS CodeDeploy's Canary deployment strategy allows you to gradually shift traffic to the new version, ensuring minimal impact on performance during the deployment.
- B. Correct.
Correct: Amazon CloudWatch alarms can monitor performance metrics and trigger automatic rollbacks if defined thresholds are breached, maintaining application health.
- C. Incorrect.
Incorrect: Manual testing is not a scalable or reliable practice for ensuring performance in a CI/CD pipeline.
- D. Correct.
Correct: Incorporating automated performance tests in a pre-production stage helps identify potential performance issues before deploying to production.
- E. Incorrect.
Incorrect: Deploying directly to production without safeguards like Canary deployments or automated tests is risky and not aligned with best practices for minimizing performance degradation.