CT_HW3

Computer Technology Course - Homework 3

---

Distributed In-Memory Database

A distributed in-memory key-value store written in Go. The system follows a single-leader architecture and is designed to be scalable, modular, and easy to deploy using Docker.

Features

• In-memory key-value storage with distributed architecture.
• Single-leader replication model.
• A central controller manages node and partition assignment.
• Key partitioning is determined by Hash(key) % number_of_partitions.
• Leader replicas use LSM trees to persist changes and initialize new replicas.
• Load balancers use the partition and node information to route requests.
• Clients can send requests and manage the cluster by adding or removing nodes and partitions.
• Dockerized deployment.

Project Structure

• Controller: Assigns keys to partitions and manages the overall cluster configuration.
• Nodes (Replicas): Store and manage data for specific partitions; leaders persist logs using LSM trees.
• Load Balancer: Uses metadata to forward requests to the appropriate nodes.
• Client: Interfaces with the system to send key-value operations and manage nodes/partitions.

Running with Docker

Run in project's root:

docker compose --profile "*" build
docker compose up

Contributors

• Mohammad Barkatin
• Mehdi Taheri
• Parisa Sadat Mousavi