Which metric can an Azure Monitor autoscale rule on an App Service plan use to add instances when the average load across them climbs?
Show answer and explanation
Answer: A
CPU percentage is a plan-level metric that Azure Monitor autoscale can evaluate, so a rule can add an instance when average CPU across the plan crosses a threshold. Memory percentage, HTTP queue length, and data in and out are other built-in options. The other choices are not metrics that autoscale uses to set instance counts. Error counts come from application telemetry and are not part of the plan autoscale engine.
Why the other options are wrong
- B. Error response counts are captured by application telemetry rather than the plan's built-in autoscale metrics, so this rule cannot be set from them by default.
- C. The age of the last deployment says nothing about current load, and Azure Monitor autoscale does not evaluate deployment timestamps when scaling out.
- D. Billing totals are cost data and are not a performance signal, so an autoscale rule cannot use them to decide when additional instances are needed.