AZ-400 Question 165
Select 2You manage a mission-critical web application running on Azure App Service with multiple instances behind a load balancer. Your team needs to deploy a new release with minimal downtime and the ability to roll back quickly if problems arise. Which two deployment approaches would best help achieve these goals?
- A
Deploy the new version to a staging slot, validate it, and perform a swap to production once testing is complete.
- B
Use a rolling deployment strategy through your Azure DevOps pipeline to update one instance at a time behind the load balancer.
- C
Perform a direct deployment to the production slot during off-peak hours to minimize user impact.
- D
Increase the App Service plan to a higher tier for extra instances and then deploy for faster release times.
- E
Completely tear down the production slot and rebuild it to ensure no residual configuration from the old deployment.
Show answer and explanation
Correct answers: A, B
Explanation
In Azure App Service, deployment slots let you perform a VIP swap from a validated environment to production, minimizing downtime and enabling quick rollbacks. For applications running behind a load balancer, rolling deployments ensure only a portion of instances are updated at a time. These approaches align with best practices for reducing impact on end users. For more information, refer to Microsoft Docs: https://learn.microsoft.com/azure/app-service/deploy-staging-slots.
- A. Correct.
Option 1: CORRECT. Using a staging slot allows you to deploy and test the new version in isolation. Once you verify everything works as expected, you can swap the staging slot with the production slot. This approach helps minimize downtime and provides an easy rollback if needed.
- B. Correct.
Option 2: CORRECT. A rolling deployment strategy updates instances behind the load balancer one at a time. As each instance is successfully updated and healthy, traffic is shifted to it, minimizing downtime and allowing for a quick rollback if something goes wrong.
- C. Incorrect.
Option 3: INCORRECT. Deploying directly to production during off-peak hours might reduce user impact, but it doesn't guarantee minimal downtime. If the deployment fails, reverting is more disruptive, and there can still be noticeable downtime for some users.
- D. Incorrect.
Option 4: INCORRECT. Increasing the tier may temporarily improve performance capacity, but it doesn't inherently prevent downtime. You still risk a brief outage or an all-or-nothing rollout, making rollbacks more complicated.
- E. Incorrect.
Option 5: INCORRECT. Completely rebuilding the production slot is time-consuming, introduces unnecessary risk, and almost guarantees more downtime than a well-planned slot swap or rolling update.