Skip to content

BN Engine Actions

Technical reference for BN Engine analysis actions. Use these for validation, diagnostics, and capacity planning.

Summary

Quick system-level availability snapshot and component overview.

Details

  • Outputs: per‑component availability, expected downtime, system availability
  • Purpose: An at‑a‑glance table: availability, expected downtime, and a system‑level snapshot
  • When: fast checks, model validation, executive summaries
  • Action: run to get a one‑page table for triage

Run Simulation

Detailed analysis with table and graph views showing exact BN calculations.

Details

  • Table view: numerical results, downtime calc, sortable, CSV export
  • Graph view: availability overlay on diagram, SVG export
  • Purpose: Deterministic BN analysis with full numerical outputs and visual exports
  • Use: in‑depth reports, reproducible technical results

Impact Analysis

Identifies components with the highest impact on system availability/reliability.

Mathematical Foundation: For each component, the impact score is calculated as:

Impact(component) = P(Root=UP) - P(Root=UP | component=DOWN)

Where: - P(Root=UP) is the normal system availability - P(Root=UP | component=DOWN) is system availability when forcing the component to be DOWN - Higher impact scores indicate more critical components

Details

  • Table view: numerical results, downtime calc, sortable, CSV export
  • Graph view: impact overlay on diagram, SVG export
  • Purpose: Ranks components by their contribution to overall unavailability/reliability loss
  • Output: Produces a ranked list of top contributors with contribution percentages for prioritization
  • Use: prioritize fixes, capacity investments, and mitigation priorities

Root Cause (Posterior)

Identifies most likely failed components given evidence of system failure.

Mathematical Foundation: Uses Bayes' theorem to calculate posterior probabilities:

P(component=DOWN | Root=DOWN) = P(component=DOWN, Root=DOWN) / P(Root=DOWN)

The analysis ranks components by their likelihood of being the root cause given the observed system state.

Details

  • Method: Bayesian inference on observed outage evidence to determine which components are most likely to have caused observed system failures
  • Outputs: Ranked list of components by failure probability given evidence for focusing troubleshooting efforts and guiding diagnostic procedures

Sensitivity Explorer

Interactive what-if scenarios for testing component availability impacts.

Mathematical Foundation: Calculates conditional probabilities for target nodes given evidence:

P(target_nodes=UP | evidence_conditions)

Allows forcing single or multiple component states and comparing availability deltas against baseline scenarios.

Details

  • Purpose: Test single or multiple component states to measure availability impact and validate redundancy assumptions
  • Output: Availability deltas and conditional probabilities for different evidence scenarios
  • Use: Design testing, scenario planning, redundancy validation, and design trade-offs

Choosing an Action

Guide

  • Summary: fast overview; use for triage
  • Run simulation: full numeric report; use for documentation and analysis
  • Impact analysis: prioritize investments and fixes
  • Root cause: incident triage when system is down
  • Sensitivity: design testing and scenario planning

Next Steps