Exam AWS Certified Developer - Associate DVA-C02 topic 1 question 128 discussion

Option C is the correct solution that meets the requirements. Performing an immutable update to deploy the new application code to new EC2 instances and serving traffic to the new instances after they pass health checks will ensure zero downtime for the application. Option A would work but cloning the production environment to a different platform version will result in a longer deployment time and can impact the cost of the environment.

Not C: While an immutable update can ensure zero downtime during the deployment process, it doesn't account for updating the Elastic Beanstalk platform version.

Option A creates a completely separate environment with the newer Node.js version, deploys the new code there, and allows for testing without affecting the production environment. Once verified, you simply swap the URLs

A) Eliminated - This solution meets the requirement of zero downtime but may have additional resource and cost overhead. B) Eliminated - All-at-once deployment means that all instances are updated at the same time, which would result in downtime during the deployment. C) Correct - Immutable updates ensure that new EC2 instances are created, and only healthy instances will serve traffic. The old EC2 instances are not affected until the new instances are confirmed to be healthy and live. D) Eliminated - Rolling deployments help ensure that some EC2 instances are always running the old code and serving traffic while others are being updated.

Option A is Correct not C: By cloning the production environment to a different platform version, you create a separate environment where you can safely deploy and test the new application code and platform version without affecting the live production environment.• Option C (Immutable update): While immutable updates ensure zero downtime by deploying to new instances, they do not address the need to update the Elastic Beanstalk platform version. Additionally, this approach can be more resource-intensive and costly.

Ans is c

C is the correct answer.

It is A: https://docs.amazonaws.cn/en_us/elasticbeanstalk/latest/dg/using-features.platform.upgrade.html#using-features.platform.upgrade.config

This question must be true for 2 options because C & D are both correct

https://docs.amazonaws.cn/en_us/elasticbeanstalk/latest/dg/using-features.rolling-version-deploy.html

The solutions that best meet the requirements for zero downtime are: A. Clone the production environment to a different platform version. Deploy the new application code, and test it. Swap the environment URLs upon verification. C. Perform an immutable update to deploy the new application code to new EC2 instances. Serve traffic to the new instances after they pass health checks. Both options A and C provide robust strategies for deploying updates with zero downtime, allowing for thorough testing in an isolated environment before directing production traffic to the new setup.

Not C: tt doesn't account for updating the Elastic Beanstalk platform version. This would affect both the live and test environments. Its also best practise to have 2 seperate environments for production and test and there is no mention of cost optimisation here.

ChatGPT:A

A & C both works for given scenario but C does it more feasibly for Elastic Beanstalk with zero downtime.

Key terminology in question is "Test". So it should be immutable for quick rollback in case of test not working.

Explanation: Immutable Update with Elastic Beanstalk: With an immutable update, Elastic Beanstalk provisions new instances with the updated code while keeping the existing instances running. The traffic is shifted gradually to the new instances after they pass health checks, ensuring that there is no downtime during the deployment. If any issue arises during the deployment, traffic is still being served by the existing instances.

Screenshot of Step 4 of Method 1 in the link: https://docs.amazonaws.cn/en_us/elasticbeanstalk/latest/dg/using-features.platform.upgrade.html#using-features.platform.upgrade.config "...your application is unavailable during the update. To keep at least one instance in service during the update, enable rolling updates"