Examkin

Soru

Zorluk: ZorScale Azure App Service Web Apps

A cloud architect is designing a performance optimization strategy for a critical REST API hosted on an Azure App Service. The API currently runs on a Basic (B1B1) App Service plan. During seasonal promotions, the API experiences high CPU utilization, and the team wants to implement a dynamic scaling solution. The configuration must automatically add instances when average CPU exceeds 80%80\%, remove instances during low-load periods, and prevent autoscale flapping. Which two actions should be performed to meet these requirements? (Select two.)

  1. Change the App Service plan pricing tier to Standard (S1S1) or higher.Cevap
  2. Ensure the scale-in CPU threshold is less than the scale-out CPU threshold multiplied by the ratio of current instances to post-scale-out instances.Cevap
  3. C
    Configure autoscale rules directly within the current Basic (B1B1) tier.
  4. D
    Set the scale-in CPU threshold to a value equal to or greater than the scale-out CPU threshold.

Cevap

Change the App Service plan pricing tier to Standard (S1S1) or higher, and ensure the scale-in CPU threshold is less than the scale-out CPU threshold multiplied by the ratio of current instances to post-scale-out instances.
To enable autoscale rules, the App Service plan must be upgraded to the Standard tier or higher because the Basic tier only supports manual scaling. To prevent autoscale flapping, the scale-in CPU threshold must be set below the expected load drop that occurs when the plan scales out. This drop is calculated by multiplying the scale-out threshold by the ratio of current instances to post-scale-out instances.

Adım Adım Çözüm

1
Identify the minimum pricing tier required for autoscale.
The App Service plan must be upgraded to Standard (S1S1) or higher, as the Basic (B1B1) tier only supports manual scaling.
Autoscale rules cannot be configured on the Basic (B1B1) tier.
2
Analyze the mathematical condition to prevent flapping during scale-in and scale-out events.
The scale-in threshold must be set below the post-scale-out load (Lin<Lout×NcurrentNnewL_{in} < L_{out} \times \frac{N_{current}}{N_{new}}).
If the scale-in threshold is too high, the redistribution of load among the new instances will immediately trigger a scale-in action, leading to a loop of scaling up and down.

Anahtar Kavram

Autoscale rules and tier support in Azure App Service
Bu soruyu puanla

Konular

Scale Azure App Service Web Apps

Bu konudan daha fazla soru