The current Openshift and Universal Monitor Agent(UMA) monitors do not provide the below useful information required for capacity planning
1. Raw Metrics
- CPU Utilization (mCore) : Total amount of mCore used by the Container.
- CPU Request (mCore) : Total amount of mCore requested by the Container. Metrics are rolled upto Pod, Project and Node level
- Memory Request (Bytes) : Total amount of memory requested in bytes by the container
- Memory Usage (Bytes) : Total amount of memory used in bytes by the container
2. Aggregate them at the following level
- Pods : Sum of individual metrics of all underneath containers
- Namespace/Project : Sum of individual metrics of all underneath pods running on a namespace/project
- Node : : Sum of individual metrics of all underneath pods running on a node
3.Deviation Metrics
a)Calculated with the following formula : 100 * (Utilization – Requested )/Requested
b) 2 metrics:
- Memory Request Deviation
- CPU Request Deviation
c)Aggregated at Pod, Project and Node level
c) Taken into consideration for Aggregated Raw Metrics collected at each level
d) Shows 0 value if Requested = 0
- APM Openshift Monitor for onpremise and SaaS
- APM Universal Monitor Agent (UMA) for Openshift in SaaS