The exhibit shows that you have an autoscale rule configured for your App Service app named App1. The rule is based on the memory percentage metric, which measures the average amount of memory used by all the instances of your app. The rule has the following settings:
Scale out action: Add 1 instance when the memory percentage is greater than or equal to 80% for a duration of 10 minutes.
Scale in action: Remove 1 instance when the memory percentage is less than or equal to 60% for a duration of 10 minutes.
Instance limits: The minimum number of instances is 2, and the maximum number of instances is 5.
According to the question, during a 30-minute period, App1 uses 60% of the available memory. This means that the scale in action is triggered, but not the scale out action. Therefore, one instance is removed from App1 every 10 minutes, until the minimum number of instances is reached.
Since App1 initially has two running instances, after the first 10 minutes, one instance is removed and App1 has one instance left. However, since the minimum number of instances is set to 2, another instance is added back to App1 to meet the minimum requirement. Therefore, after the first 10 minutes, App1 still has two instances.
After the second 10 minutes, the same process repeats. One instance is removed due to the scale in action, and another instance is added back due to the minimum requirement. Therefore, after the second 10 minutes, App1 still has two instances.
After the third 10 minutes, there is no change in the number of instances, because App1 already has the minimum number of instances. Therefore, after the third 10 minutes, App1 still has two instances.
Therefore, during the 30-minute period, App1 never has more than two instances running at any given time. However, since one instance is removed and added back every 10 minutes, there are four different instances that are used by App1 during the period. Hence, the maximum number of instances for App1 during the period is four.
