slayer

South of Heaven, North of Five Nines.

slayer is a reliability modeling and probabilistic risk analysis suite designed for infrastructure architects and SREs. It moves beyond simple calculators to model complex, multi-layered system availability, disaster recovery profiles, and yearly risk distributions.

Live version: https://mrueg.github.io/slayer/

• Hierarchical Modeling: Build complex dependency trees with nested groups and components.
• K-out-of-N Redundancy: Model exact success probability for nodes requiring minimum capacity (e.g., a 16-node blade chassis requiring 12 nodes UP).
• Failover Switching: Parallel configurations include a "Failover %" to model the reliability of the switch/load-balancer mechanism itself.
• MTTR & Frequency: Derives system-wide Mean Time To Recovery using frequency-weighted averages of individual component restoration times.
• RTO & RPO Modeling: Mathematically verifies Disaster Recovery goals by aggregating Recovery Time Objectives and Recovery Point Objectives across the tree.

• Monte Carlo Simulation: Executes 10,000 yearly iterations using Poisson arrival rates for incidents and Exponential distributions for recovery durations.
• Distribution Histogram: Visualizes the "long-tail" risk of your architecture, showing the probability of "Bad Year" (P95) and "Calamity" (P99) scenarios.
• Sensitivity-Based Bottlenecks: Automatically identifies the "System Bottleneck"—the specific node where improvements would yield the highest gain in total system health.

• Chaos Mode: An interactive failure injection environment.
• Granular Failure Magnitude: For redundant components, specify exactly how many replicas to "Kill" to observe system degradation.
• Blast Radius Heatmap: A visual overlay that colors nodes based on the percentage of total system SLA lost if they fail.
• Real-time Error Budgeting: Tracks consumed downtime against allowed duration for specific periods (Day/Month/Year).

• Infrastructure Topology: Toggle between a hierarchical list editor and a visual infrastructure diagram (Horizontal/Vertical).
• Technical Annotations: Persistent technical notes and assumptions visible directly on the architecture nodes.
• Cloud SLA Catalog: Searchable database of published SLAs from AWS, Azure, GCP, and Cloudflare to pre-populate models.

slayer is built with Next.js 15, Tailwind CSS 4, and Lucide React. It utilizes several sophisticated mathematical models:

• Probability: Exact success probability via dynamic programming for non-identical component SLAs in K-of-N groups.
• Stochastics: Poisson arrival rate modeling for incident frequency.
• Simulation: 10,000-run Monte Carlo engine with debounced real-time updates.

npm install

npm run dev

slayer uses a "Shareable State" model. Your entire infrastructure design is encoded into a Base64 URL hash. You can share your model simply by copying the link—no database required. It also supports local JSON Export/Import for file-based configuration management.

Apache-2.0