This article provides a high level sizing reference on monitoring OpenShift with APM Infrastructure Agent. As a general rule, the resource requirement is often directly proportional to the amount of monitoring required.
What are the basic CPU and Memory requirements of the CA APM Infrastructure Agent?
CA Application Performance Management 10.7 OpenShift v3.x
The resources (CPU and memory) consumption for monitoring OpenShift with CA APM Infrastructure Agent depends on what is being monitored.
The Cluster monitoring is done via the Deployment Pod and node monitoring is done via Daemonset Pod. For CA APM Infrastructure Agent-based monitoring, it depends on the amount of Kubernetes Objects you are monitoring.
Cluster monitoring Agent will need about 512 MB of memory and 0.5 CPU ( assuming 1000 pods in the cluster ) where as node monitoring agent can be configured with 0.3CPU with 256 MB of memory( 100 pods per node ).