ros-ocp-backend

The Resource Optimization Service (ROS) backend for OpenShift provides rightsizing recommendations for containers, deployments, and jobs. It integrates with Kruize as the optimization engine.

What It Does

ROS analyzes resource usage data collected by the koku-metrics-operator and produces recommendations to help teams right-size their workloads:

Cost-optimized recommendations minimize resource spend
Performance-optimized recommendations ensure workloads have sufficient headroom

Recommendations cover CPU and memory requests and limits for containers.

Architecture

ROS sits alongside Koku and shares the same metering data pipeline:

    flowchart LR
    Operator[koku-metrics-operator] -->|metrics| Koku[Koku Pipeline]
    Koku --> ROS[ros-ocp-backend]
    ROS --> Kruize[Kruize Engine]
    Kruize --> ROS
    ROS --> API[ROS API]
    API --> UI[koku-ui ROS]

The ROS backend aggregates data, sends it to Kruize for analysis, and exposes the resulting recommendations through its own API — which the ROS UI in koku-ui consumes.

ros-ocp-backend

ros-ocp-backend

What It Does

Architecture

Further Reading