SLOpilot On-Prem User Guide

SLOpilot is a Kubernetes resource optimization platform that analyzes real-time cluster metrics to deliver data-driven CPU and memory recommendations. Deployed into your existing cluster, it helps platform teams reduce infrastructure costs while maintaining application reliability.

Architecture

graph TD
    K8S[Kubernetes API<br/>workload discovery]
    PROM[Bundled Prometheus<br/>+ kube-state-metrics]
    SLO[SLOpilot Pod<br/>Application Server + Embedded Web UI]
    LIC[License Server<br/>external SaaS]

    K8S -->|scrapes metrics| PROM
    PROM -->|metric queries| SLO
    K8S -->|workload discovery| SLO
    SLO -->|license validation| LIC

Key Capabilities

• Automated CPU and memory recommendations with confidence scoring — recommendations grow more precise as metric history accumulates.
• Conservative approach — built-in safety margins prevent resource contention and application disruption.
• HPA and VPA awareness — adapts its output when horizontal or vertical autoscalers are already managing a workload.
• PDF report generation — shareable compliance-ready reports for stakeholder communication.
• Bundled monitoring stack — no external Prometheus installation required; SLOpilot ships its own.

How It Works

SLOpilot deploys into your Kubernetes cluster alongside a bundled Prometheus instance that begins collecting CPU and memory metrics immediately. After statistically meaningful data collection (approximately one to two weeks, depending on the workload), the analysis engine produces recommendations with increasing confidence. No application changes, sidecars, or code modifications are required.

Next Steps

• Prerequisites — verify your environment before installing.
• Installation — run the installer and deploy SLOpilot into your cluster.