1Z0-1072-25 Question 20
Select 2Your marketing team is about to launch a campaign that could sporadically spike web traffic on an application running in an Oracle Cloud Infrastructure instance pool. You want to configure threshold-based autoscaling to scale out quickly once CPU usage hits 70% for at least 5 minutes, and to scale in if usage remains below 30% for 10 minutes. You also want to prevent rapid, repetitive scaling from brief traffic flurries. Which two configurations are required to achieve these goals?
- A
Define a threshold-based policy that triggers scale-out if average CPU usage remains above 70% for a specific measurement period, and triggers scale-in if average CPU usage stays below 30%.
- B
Configure the autoscaling policy to instantly launch or terminate instances whenever CPU usage crosses either threshold, regardless of duration.
- C
Implement a schedule-based autoscaling policy that scales out at a fixed time each day and scales in at a fixed time each night.
- D
Set a cooldown period to avoid immediate re-scaling during short-lived usage fluctuations after scaling actions occur.
- E
Assign a fixed number of additional instances to start up whenever the threshold is reached, rather than specifying a maximum or minimum size.
Show answer and explanation
Correct answers: A, D
Explanation
In Oracle Cloud Infrastructure, threshold-based autoscaling policies let you specify metrics (CPU usage) and measurement windows to ensure sustained usage before triggering scale actions. Combining thresholds (70% to scale out, 30% to scale in) with a proper cooldown period is essential to avoid rapid scaling churn. For more details, refer to OCI documentation on instance pool autoscaling and best practices for threshold-based policies.
- A. Correct.
Correct. A threshold-based policy requires specifying the metric thresholds (70% for scale-out, 30% for scale-in) and a measurement window. This ensures the policy only triggers when usage meets or exceeds those thresholds for the designated duration rather than reacting to transient spikes.
- B. Incorrect.
Incorrect. Immediately launching or terminating instances whenever CPU usage crosses a threshold can lead to frequent, destabilizing scaling actions for short-lived spikes. A duration or measurement period is needed to confirm sustained usage before scaling.
- C. Incorrect.
Incorrect. Schedule-based autoscaling triggers at set times, regardless of actual CPU usage. In sporadic traffic scenarios, threshold-based autoscaling is more appropriate to handle unpredictable loads.
- D. Correct.
Correct. A cooldown period prevents the autoscaling service from continuously resizing the instance pool if the CPU usage briefly dips below or spikes above thresholds. It stops frequent back-and-forth scaling by waiting a specified amount of time before another scale action.
- E. Incorrect.
Incorrect. Rigidly assigning a fixed number of additional instances on threshold breaches removes flexibility. A preferred approach is to define minimum and maximum sizes and let autoscaling scale within that range based on usage.