A Deployment has a HorizontalPodAutoscaler targeting 70 percent average CPU. Observed utilization climbs to 90 percent. What does the HPA do?
Show answer and explanation
Answer: A
The Horizontal Pod Autoscaler compares observed metrics with the target and adjusts the replica count of the workload it controls. When average CPU is above the target, it adds replicas so the load spreads across more pods, within the configured minimum and maximum. It does not terminate pods to free capacity, change machine types, or roll back revisions. Those actions belong to other mechanisms.
Why the other options are wrong
- B. Removing pods happens when metrics fall below the target, and only after a stabilization period. Load above target triggers scale-up, not termination.
- C. Machine type changes are not an HPA function. Node resizing is handled at the node-pool level, not by pod autoscaling.
- D. Rollbacks are a deployment action triggered by people or pipelines. HPA does not revert revisions when CPU is high.