Platform Health Assessment

[digitalrisedorset]

Summary

Introduce a Platform Health Assessment component responsible for interpreting platform observations into deterministic health assessments.

The Health Assessment consumes platform snapshots produced by Platform Signals and evaluates the operational state of the platform using reusable health rules.

This feature introduces the decision layer between platform observation and future workload adaptation.

Objectives

• Introduce a Platform Health Assessment model.
• Evaluate platform snapshots using deterministic rules.
• Produce reusable health assessments.
• Record health assessments through OpenTelemetry.
• Remain independent from Platform Signals and workload implementations.

Responsibilities

The Platform Health Assessment is responsible for:

• Evaluating load average.
• Evaluating CPU utilisation.
• Evaluating memory pressure.
• Evaluating Redis availability.
• Evaluating Varnish cache efficiency.
• Producing an overall platform assessment.

The Platform Health Assessment does not:

• Collect platform signals.
• Execute remediation.
• Pause workloads.
• Modify crawler behaviour.

Example

{
  "healthy": true,
  "severity": "warning",
  "score": 82,
  "observations": [
    "Load average exceeds logical core count.",
    "Varnish cache efficiency continues to improve."
  ],
  "recommendation": "Continue cache warming."
}

Design Principles

• Deterministic rule evaluation.
• Platform-agnostic assessment.
• Independent from Platform Signals.
• Independent from workload implementations.
• Reusable by future Health Engine features.

Future Enhancements

Future iterations may consume the Platform Health Assessment to:

• Pause cache warming.
• Reduce crawler concurrency.
• Resume workloads.
• Trigger remediation workflows.
• Drive autonomous Health Engine decisions.