⚙️ TECHNICAL ARCHITECTURE & ENGINE

How Our Compiler Works

Understand the inner workings of Our Compiler. From client-side Monaco Editor interactions to cloud container sandboxing and sub-2-second execution output delivery.

🔄 Code Execution Flowchart

When you click Run Code in Our Compiler, your source program passes through five distinct technical layers to ensure fast, isolated, and safe execution:

STEP 1: FRONTEND CAPTURE

Monaco Editor

The editor captures source code, language configuration, active tab state, and interactive custom inputs (stdin).

STEP 2: ENCRYPTED DISPATCH

TLS 1.3 Payload

The code payload is securely serialized into JSON and sent over an encrypted HTTPS connection to execution API endpoints.

STEP 3: SANDBOX CONTAINER

Docker Isolation

An ephemeral Docker container initializes with strict resource constraints (512MB RAM, 10s CPU limit, no root access).

STEP 4: COMPILATION & EXECUTION

Native Toolchain

The runtime toolchain (e.g. GCC 11, JDK 17, Python 3.10) compiles and executes the program against stdin buffers.

STEP 5: STREAMED OUTPUT

Console Display

Standard output (stdout), compilation errors (stderr), and execution status (exit code 0/1) are returned to the user console.

🛡️ Container Security & Isolation

Security is the highest priority when building a public online compiler. Allowing arbitrary user code execution means preventing malicious system calls, infinite loops, and resource exhaustion. Our Compiler handles security via:

CPU & Time Limits: Programs execution is limited to 10 seconds. Infinite loops are terminated automatically.
Memory Constraints: Each sandbox container is capped at 512MB RAM.
Disabled Network Access: Execution containers operate in offline network modes to prevent unauthorized outbound requests.
Ephemeral Disk Storage: Created temporary files are destroyed immediately after execution finishes.

⚡ High-Performance Latency Optimization

To deliver execution times under 2 seconds, Our Compiler utilizes pre-warmed container pools. Instead of spinning up a fresh container on every request, idle container runtimes stand ready to compile code instantly, eliminating cold-start delays.

Whether you write Python script, C++ template code, or Java object-oriented classes, your output appears almost instantly.