Overview

Maybets is a high-performance Go application designed for real-time processing of large volumes of betting transactions while providing analytical insights. It efficiently ingests transaction data, stores it, and exposes APIs for retrieving key betting analytics.

Features

• Data Ingestion: Accepts betting transactions in JSON format from a file (bets.json) or an API endpoint.
• Processing & Storage: Leverages Go's in-memory data structures and SQLite for efficient transaction handling.
• Analytics APIs: Provides insights into user betting statistics and detects anomalies.
• Performance Optimization: Uses goroutines for concurrent processing, ensuring a throughput of at least 10,000 bets per second.
• CLI Support: Includes a command-line interface for batch processing.

Approach

The approach below was adopted when implementing the solution

Batch Inserts

Instead of inserting bets individually, transactions are grouped into batches (e.g., 1,000 bets per batch) and inserted in a single transaction. This significantly reduces database writes and improves performance.

Connection Pooling

A connection pool manages database connections efficiently, ensuring the system can handle multiple concurrent requests without overwhelming SQLite.

Concurrent Processing

Goroutines process bets concurrently, enabling the system to handle multiple batches of bets simultaneously and improving throughput.

Indexed Queries

Frequently queried columns (e.g., user_id) are indexed to optimize SQL query performance and speed up retrieval times.

Trade-offs

Increased Memory Usage

Batch inserts require temporarily storing transactions in memory before writing to the database, leading to higher memory consumption, especially with large datasets.

SQLite Limitations

SQLite is lightweight and not designed for extremely high write loads. While batching and connection pooling mitigate some limitations, SQLite may still struggle under very high transaction volumes.

Setup & Installation

Prerequisites

• Go 1.20+
• Redis (MANDATORY)

Installation & Running

1. Clone the Repository

git clone [email protected]:KathurimaKimathi/maybets.git
cd maybets

2. Install Dependencies

go mod tidy

3. Start the Redis Server (if caching is enabled)

Ensure Redis is installed and running:

redis-server

4. Set Up Environment Variables

export ENVIRONMENT="LOCAL"
export REDIS_URL="redis://localhost:6379/0"
export JAEGER_ENDPOINT="localhost:4318"
export PORT="8080"
export SQLITE_URL="/path/to/your/sqlite/file"

5. Run the Server

Method 1: Using CLI

cd cmd
go run cmd.go runserver

Method 2: Direct Execution

go run server.go

6. Generate and Load Test Data

Generate test betting data:

cd cmd
go run cmd.go generate --betdata 10000 bets.json

Load test data into SQLite:

go run cmd.go process bets.json

API Reference

Each betting transaction follows this JSON structure:

{
  "bet_id": "string",
  "user_id": "string",
  "amount": "float64",
  "odds": "float64",
  "outcome": "win" | "lose",
  "timestamp": "RFC3339 format"
}

Endpoints

1. Get Total Bets by User

curl --location '<BASEURL>:<PORT>/api/v1/analytics/total_bets?user_id={user_id}'

2. Get Total Winnings by User

curl --location '<BASEURL>:<PORT>/api/v1/analytics/total_winnings?user_id={user_id}'

3. Get Top 5 Users by Betting Volume

curl --location '<BASEURL>:<PORT>/api/v1/analytics/top_users'

4. Get Users with Anomalous Betting Activity

curl --location '<BASEURL>:<PORT>/api/v1/analytics/anomalies'

Tracing (WIP)

Run the following command to start the tracing service:

docker compose up

Then, navigate to Jaeger UI to view request traces across various points in the application.

Contribution

Contributions are welcome! Feel free to submit issues, feature requests, or pull requests to improve the project.

License

This project is licensed under the MIT License.