Faultline

Break systems safely. Replay the same traffic with a fix.

2 followers

Break systems safely. Replay the same traffic with a fix.

2 followers

An interactive lab for distributed systems. Trigger outages, trace cascading failures, and replay identical traffic to test a fix. Explore service pressure, recovery, and availability targets. Free to use, with no account or API key.

Free

Launch tags:Education•Software Engineering•Developer Tools

Launch Team

Dropbox Sign APIBuild eSignatures into your app with less work.

Promoted

OpenAI

GPT-6 Astra Challenge

What became possible in your product with Astra that was not practical before?

Maker

The breakthrough was taking Faultline from an idea to a working, tested, publicly deployed resilience lab in one evening. The challenging part was keeping the simulation and interface consistent. Individual requests move through finite service pools, bounded queues, retries, timeouts and circuit breakers. The interface exposes the resulting behavior. The key feature is counterfactual replay: send identical seeded traffic through different configurations and measure whether a fix helps. Users can trace requests, adjust outages, inspect service pressure, compare availability and latency, and export the evidence. Eight automated checks cover deterministic replay, request conservation, bounded capacity and recovery. This made a complete interactive learning tool practical within a short build session. The deployed simulation runs locally in the browser without an AI API.

Report

2mo ago

Maker

📌

Hi Product Hunt 👋 A payment service stops responding. Why does browsing slow down too? I built Faultline to make cascading failures visible and give developers somewhere safe to experiment. Start with a simulated storefront. Trigger an outage, watch queues build, and follow individual requests through timeouts and retries. Pause or rewind whenever something surprises you. Then change a safeguard and replay the exact same traffic. Does a circuit breaker help? Do extra replicas solve the bottleneck? What happens to the requests that still fail? Faultline includes: • Four scenarios: retry storms, cold caches, traffic spikes, and slow database queries. • Adjustable traffic, outage timing, and safeguards. • A service pressure timeline with moments you can inspect. • Availability targets, error budgets, and recovery analysis. • Comparisons of success rate, response time, and retry amplification. • Reproducible experiment links and downloadable evidence. Everything runs in your browser, with no account or API key. It’s an educational simulation for learning and explaining system behavior. Try “The retry storm” → “Jump to fault” → “Compare a fix,” then explore the Analysis tab. Which failure mode would you like to investigate next?

Report

2mo ago

Reviews

OpenAI

GPT-6 Astra Challenge

What became possible in your product with Astra that was not practical before?

Maker

Report

2mo ago

Maker

📌

Report

2mo ago