The new ‘Citrix Intelligent Autoscale’ in CVAD 2511 is configured to save money. Instead, your hypervisor logs are full of machines powering on and off every 15 minutes, and users are complaining about slow session launches.

State: Flapping, Autoscale Metric: Performance Target Mismatch, VDA: 2511.

The data-driven performance target is set too aggressively, causing the capacity buffer to oscillate when the logon rate exceeds the hypervisor’s power-on latency.

Switch from “Performance Target” to a static “Percentage of Machines” buffer (10%) and increase the “Power-off Delay” to 30 minutes.

A surgical script fixes the flapping, but it doesn’t fix the architectural root causing the waste. [Stop the firefighting]