[ResourceMonitoring] Abnormal numbers for cpu.time and memory.usage #6815
Description
Description
We are using latest Microsoft.Extensions.Diagnostics.ResourceMonitoring (9.8.0) and recently added container.cpu.time and container.memory.usage metrics. I am a little bit confused by numbers I see. They are not even close to numbers from k8s.
Our configuration:
- There are sidecar containers in pod
- cpu.limit < 1 (in case you use it for scaling internally)
Reproduction Steps
Collect new metrics
Expected behavior
Numbers similar to metrics from k8s
Actual behavior
Unrealistic numbers
Regression?
No response
Known Workarounds
No response
Configuration
No response
Other information
No response