I would like to know to know exactly how the metric "CPU usage" is calculated in dynaTrace, particularly in the context of using processors with adjustable frequency depending for example on load and temperature - think of turbo boost, speed steps, etc..
If i want to respect Microsoft's recommendation not to exceed 60% CPU load on their IIS web servers and are not currently able to because measures from dT and windows don't match in their load testing scenarios.
Can you please explain clearly how CPU frequency is calculated in dT ?
Answer by Andreas G. ·
Measures in general are captured/calculated in a 10 second interval. For CPU or other Host/Process/Infrastructure measures we typically use OS APIs to query that data, or use Performance Counters from the OS (e.g: Windows Performance counters).
As for CPU Usage which is shown in the Host Health Dashboards - please have a look at Host Health Monitoring