Examkin

Question

Difficulty: MediumScale Azure App Service Web Apps

You manage a web application named LogiRoute that runs on a Standard (S2) App Service plan. The plan currently has 2 instances. You configure an autoscale scale-out rule to increase the instance count by 1 when the average CPU percentage exceeds 80%80\%. During a peak period, the average CPU utilization across the 2 instances reaches 84%84\%, triggering a scale-out event. You need to configure a scale-in rule to decrease the instance count by 1 when the load decreases, ensuring that the scale-in action does not immediately trigger another scale-out (preventing autoscale flapping). Assuming the total workload remains constant immediately after scaling, what is the maximum CPU percentage threshold you should set for the scale-in rule?

  1. A CPU percentage threshold of 50%50\% on the Standard (S2) App Service planAnswer
  2. B
    A CPU percentage threshold of 60%60\% on the Standard (S2) App Service plan
  3. C
    A CPU percentage threshold of 70%70\% on the Standard (S2) App Service plan
  4. D
    A CPU percentage threshold of 50%50\% on a Basic (B1) App Service plan

Answer

A CPU percentage threshold of 50%50\% on the Standard (S2) App Service plan
The correct option is correct because the total workload of 168%168\% is divided across 3 instances after scaling out, resulting in 56%56\% CPU load per instance. To prevent immediate scale-in (flapping), the scale-in threshold must be set below 56%56\%. A threshold of 50%50\% on a Standard (S2) plan is valid and prevents flapping.

Step-by-Step Solution

1
Calculate the total CPU workload before scaling.
Total CPU workload is 2 instances×84%=168%2 \text{ instances} \times 84\% = 168\%.
Determining the total aggregate CPU utilization helps predict the load per instance when distributed across more instances.
2
Calculate the expected CPU utilization per instance after scaling out to 3 instances.
Expected CPU utilization per instance is 168%/3=56%168\% / 3 = 56\%.
This shows the immediate CPU level that the autoscale engine will evaluate after the scale-out event completes.
3
Determine the maximum scale-in threshold to prevent flapping.
The scale-in threshold must be strictly less than 56%56\%. Therefore, 50%50\% is the correct and safe threshold.
If the scale-in threshold is set to a value higher than 56%56\% (such as 60%60\% or 70%70\%), the autoscale engine will detect that the CPU (56%56\%) is below the threshold and immediately scale in, causing a continuous loop of scaling out and in (flapping).

Key Concept

Autoscale flapping prevention and pricing tier constraints
Rate this question

Topics

Scale Azure App Service Web Apps

More questions from this topic