MONGODB_DATA_STRUCTURE.md

MongoDB Data Storage Structure

This document describes exactly what data is stored in MongoDB when you use the Image Colorization API.

Database: colorization_db

Collection 1: api_calls

Stores every API endpoint call with timestamp.

Data Structure:

{
    "_id": ObjectId("..."),                    // MongoDB auto-generated ID
    "endpoint": "/health",                      // API endpoint path
    "method": "GET",                            // HTTP method (GET, POST)
    "status_code": 200,                        // HTTP status code
    "timestamp": ISODate("2025-11-27T04:22:27.823Z"),  // UTC timestamp when API was called
    "request_data": {                          // Request parameters/data
        "filename": "Descratch.png",
        "content_type": "image/png"
    },
    "response_data": {                         // Response data returned
        "status": "healthy",
        "model_loaded": true
    },
    "error": null,                             // Error message if any
    "user_id": null,                           // User ID if authenticated
    "ip_address": "127.0.0.1"                 // Client IP address
}

Example for /upload endpoint:

{
    "endpoint": "/upload",
    "method": "POST",
    "status_code": 200,
    "timestamp": ISODate("2025-11-27T10:00:00.000Z"),
    "request_data": {
        "filename": "Descratch.png",
        "content_type": "image/png"
    },
    "response_data": {
        "success": true,
        "image_id": "abc123-def456-ghi789",
        "image_url": "https://.../uploads/abc123.jpg",
        "filename": "abc123.jpg"
    },
    "error": null,
    "user_id": null,
    "ip_address": "192.168.1.100"
}

Example for /colorize endpoint:

{
    "endpoint": "/colorize",
    "method": "POST",
    "status_code": 200,
    "timestamp": ISODate("2025-11-27T10:00:05.000Z"),
    "request_data": {
        "filename": "Descratch.png",
        "content_type": "image/png",
        "positive_prompt": "colorize this image",
        "negative_prompt": "low quality",
        "seed": 123,
        "num_inference_steps": 8
    },
    "response_data": {
        "success": true,
        "result_id": "xyz789-abc123-def456",
        "download_url": "https://.../results/xyz789.png",
        "api_download_url": "https://.../download/xyz789",
        "filename": "xyz789.png",
        "caption": "colorized image"
    },
    "error": null,
    "user_id": null,
    "ip_address": "192.168.1.100"
}

---

Collection 2: image_uploads

Stores information about every image upload.

Data Structure:

{
    "_id": ObjectId("..."),                    // MongoDB auto-generated ID
    "image_id": "abc123-def456-ghi789",        // Unique image identifier (UUID)
    "filename": "Descratch.png",               // Original filename
    "file_size": 245678,                       // File size in bytes
    "content_type": "image/png",               // MIME type
    "uploaded_at": ISODate("2025-11-27T10:00:00.000Z"),  // UTC timestamp
    "user_id": null,                           // User ID if authenticated
    "ip_address": "192.168.1.100"              // Client IP address
}

Example with Descratch.png:

{
    "image_id": "abc123-def456-ghi789",
    "filename": "Descratch.png",
    "file_size": 245678,
    "content_type": "image/png",
    "uploaded_at": ISODate("2025-11-27T10:00:00.000Z"),
    "user_id": null,
    "ip_address": "192.168.1.100"
}

---

Collection 3: colorizations

Stores information about every colorization request.

Data Structure:

{
    "_id": ObjectId("..."),                    // MongoDB auto-generated ID
    "result_id": "xyz789-abc123-def456",       // Unique result identifier (UUID)
    "image_id": "abc123-def456-ghi789",        // Original image identifier
    "prompt": "colorize this image",           // Text prompt used (if any)
    "model_type": "sdxl",                      // Model type: "fastai", "pytorch", "sdxl", "gan"
    "processing_time": 3.45,                  // Time taken to process in seconds
    "created_at": ISODate("2025-11-27T10:00:05.000Z"),  // UTC timestamp
    "user_id": null,                           // User ID if authenticated
    "ip_address": "192.168.1.100"              // Client IP address
}

Example with Descratch.png:

{
    "result_id": "xyz789-abc123-def456",
    "image_id": "abc123-def456-ghi789",
    "prompt": "colorize this image with vibrant natural colors",
    "model_type": "sdxl",
    "processing_time": 3.45,
    "created_at": ISODate("2025-11-27T10:00:05.000Z"),
    "user_id": null,
    "ip_address": "192.168.1.100"
}

---

What Gets Stored When You Use Descratch.png

When you upload and colorize Descratch.png, the following data will be stored:

1. API Call Logs (api_calls collection):

   • /health - Health check call
   • /upload - Image upload call with request/response data
   • /colorize - Colorization call with processing parameters
   • /download - Download result call (if you download the result)
   • /results - Public result access call (if accessed)

2. Image Upload Record (image_uploads collection):

   • Image ID (UUID)
   • Filename: "Descratch.png"
   • File size in bytes
   • Content type: "image/png"
   • Upload timestamp
   • IP address

3. Colorization Record (colorizations collection):

   • Result ID (UUID)
   • Image ID reference
   • Model type used
   • Processing time
   • Creation timestamp
   • IP address

All Data Includes Timestamps

Every record includes a UTC timestamp:

• api_calls.timestamp - When the API endpoint was called
• image_uploads.uploaded_at - When the image was uploaded
• colorizations.created_at - When the colorization was created

Querying the Data

You can query MongoDB to:

• Find all API calls in the last 24 hours
• Track which images were uploaded
• Monitor colorization processing times
• Analyze API usage patterns
• Track errors and status codes