You have a Compute Engine application that you want to autoscale when total memory usage exceeds 80%. You have installed the Cloud Monitoring agent and configured the autoscaling policy as follows: ✑ Metric identifier: agent.googleapis.com/memory/percent_used ✑ Filter: metric.label.state = 'used' ✑ Target utilization level: 80 ✑ Target type: GAUGE You observe that the application does not scale under high load. You want to resolve this. What should you do?

The autoscaler needs to understand the total "used" memory, which includes not just used but also buffered, cached, and slab memory. By default, agent.googleapis.com/memory/percentused with metric.label.state = 'used' only accounts for actively used memory, not the memory consumed by the OS for caching and buffering, which is still unavailable to applications. Expanding the filter to include these states provides a more accurate representation of the memory truly consumed by the system. Changing the Target type to DELTAPERMINUTE is incorrect because memory utilization is a point-in-time measurement (GAUGE), not a rate of change. Changing the Metric identifier to agent.googleapis.com/memory/bytesused would require recalculating the target utilization, and the original metric percentused is more direct for percentage-based autoscaling. Changing the filter to metric.label.state = 'free' and the Target utilization to 20 would trigger autoscaling when 80% of memory is free, which is the opposite of the desired behavior.