Distributed Load Balancer (dsloadbalancer)

Welcome to the Distributed Load Balancer project! This system is designed as a capstone project for UBC CPSC416 and aims to efficiently distribute and manage workloads across a network of storage and worker nodes.

Features

• Dynamic Load Distribution: The system dynamically distributes data across multiple storage nodes and assigns tasks to worker nodes, ensuring efficient load balancing.
• gRPC Communication: Utilizes gRPC for efficient and robust communication between nodes.
• Scalability: Easily scale the number of storage and worker nodes to meet the demand.
• Performance Metrics: Measures and reports the time taken for the workers to produce the final result, allowing for performance optimization.

Getting Started

Prerequisites

• Ensure you have cmake installed on your system to build the project.
• The system is built and tested on UNIX-based systems.

Building the Code

Clone the repository to your local machine. Navigate to the cloned directory and run the following commands:

$ mkdir -p cmake/build  
$ cd cmake/build  
$ cmake ../..  
$ make  
$ cd ../..  

Running the System

Execute the system using the provided shell script with the desired parameters:

./runSystem [data_size] [storage_nodes] [worker_nodes]

• data_size: The number of random key-value pairs to generate and distribute.
• storage_nodes: The number of Storage Nodes in the system.
• worker_nodes: The number of Worker Nodes in the system.

For example:

./runSystem 100000 4 3

This command runs the system with 100,000 key-value pairs, 4 Storage Nodes, and 3 Worker Nodes.

Validating Results

To ensure the correctness of the system's output, you can run the validation script:

./validation

This script compares the results produced by the distributed system against a monolithic architecture implementation.

Documentation

For a deeper understanding of the system's design and implementation, refer to the following documents:

• System Design
• Project Reflection