Cloud Architect (Cloud Server Simulator)

A real-time infrastructure management sim where you scale a cloud architecture under rising traffic, shifting workloads, and security incidents. Keep reputation high, stay profitable, and prevent your system from collapsing.

Game loop

Every second (a “tick”), the simulator:

• Generates traffic based on growth + randomness
• Applies a traffic pattern (Normal Flow / Shopping Spree / Data Ingestion / Viral Content)
• Injects attacks/events occasionally (botnet DDoS, viral spikes, cooling failure, fiber cut, investor funding)
• Routes requests through your stack (ALB → WAF → CDN → App/Worker/DB)
• Computes throughput, drops, and health decay
• Updates economics (revenue vs upkeep) and reputation based on success + security outcomes

Your job is to continuously invest in capacity and efficiency while keeping operating costs (upkeep) under control.

What you can build / upgrade

Nodes (horizontal scaling)

Provision and upgrade nodes across three layers:

• APP: handles general request processing
• WORKER: handles heavy compute (uploads, writes)
• DB: handles reads/writes/search (more expensive; critical bottleneck)

Each node has:

• Tier: T1 → T2 → T3 (higher capacity, higher upkeep)
• Health: decays faster when overloaded; can crash and require repair/reboot

Components (platform scaling)

• ALB (Load Balancer): hard caps max incoming request rate
• WAF (Firewall): blocks malicious traffic; can overload under extreme load
• CDN: serves static traffic; overflow falls back to app servers
• DB Tech: multiplies effective DB capacity (efficiency)
• Cache (Redis): reduces DB read/search demand
• Queue (SQS): buffers bursts for write/upload workloads

Operations (one-click “tasks”)

Timed boosts with cooldowns:

• Flush CDN Cache: temporary CDN capacity boost
• Optimize Indexes: temporary DB efficiency boost
• Live Security Patch: temporary WAF boost

Security modes

Switch WAF behavior:

• Standard: baseline filtering, no false positives
• High Sec: better blocking, small false positive rate (may drop legit requests)
• Panic: strongest blocking, high false positive rate

Win / lose conditions

There’s no “win screen” — the challenge is how long you can survive as traffic scales.

You lose when either:

• Reputation hits 0%
• Budget falls too far into the red (bankruptcy threshold)

Tips

• Upgrade ALB early: it’s a hard cap; if you hit it you’ll drop traffic immediately.
• Watch the bottleneck: DB overload cascades into drops and reputation loss.
• Use cache + DB tech to tame read/search-heavy patterns.
• Use queue + workers for write/upload-heavy patterns.
• Avoid running hot: overload accelerates node health decay and can cause crashes.

Tech stack

• React 19 + TypeScript
• Vite
• Tailwind CSS v4
• Recharts (traffic chart)
• Lucide (icons)

Running locally (bun)

bun install
bun run dev

Then open the URL printed by Vite (usually http://localhost:5173).

Production build

bun run build
bun run preview

Formatting / linting

bun run format
bun run lint