https://github.com/t4vexx/sockets

This project is an implementation of a simulation of the MPI_REDUCE algorithm using TCP/IP sockets in C. The main idea is to divide the sum of a set of numbers among different processes (called workers), which run in parallel, following a binary tree structure.
https://github.com/t4vexx/sockets

c make mpi-library multiprocessing parallel-computing shell-script

Last synced: 3 months ago
JSON representation

This project is an implementation of a simulation of the MPI_REDUCE algorithm using TCP/IP sockets in C. The main idea is to divide the sum of a set of numbers among different processes (called workers), which run in parallel, following a binary tree structure.

• Host: GitHub
• URL: https://github.com/t4vexx/sockets
• Owner: T4vexx
• Created: 2024-09-29T22:47:51.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-03-15T21:44:04.000Z (10 months ago)
• Last Synced: 2025-04-19T09:32:15.689Z (9 months ago)
• Topics: c, make, mpi-library, multiprocessing, parallel-computing, shell-script
• Language: C
• Homepage:
• Size: 22.5 KB
• Stars: 2
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# SOCKETS PROJECT - MPI_REDUCE() SIMULATOR

## DESCRIPTION

This project is an implementation of an MPI_REDUCE algorithm simulation using TCP/IP sockets in C. The main idea is to distribute the sum of a set of numbers across different processes (called workers) that execute in parallel, following a binary tree structure.

### Features

- Parallel Architecture: Uses a binary tree approach, where workers hierarchically perform value reduction until the final result is obtained.

- Socket Communication: Data transmission between processes is done via TCP/IP sockets, simulating point-to-point communication between workers and the manager.

- Scalability: The number of workers can be easily configured, allowing execution with different binary tree sizes for parallelism.

- Synchronization: Implements synchronization to ensure that workers operate in coordination and send data to the manager only after completing the reduction operation.

## CONFIGURATION

You must have GCC and build-essentials installed on your Linux system to run the process.

To clean executable files, run:

	

	make clean

To build the project, run:

	

	make build

	

To execute the project, run:

	

	make exec

	

After executing the processes via make exec, wait for the manager to print the final result.

![image](https://github.com/user-attachments/assets/a8db22d3-f8f9-4f6f-a949-82f390cfad22)

### Considerações

The worker initialization format is as follows: 

```Bash

	./worker  

```

In the logic used to spawn worker processes, the exec_trabalho.sh script starts all workers with the value 10.

If you want to change the values used for summation, simply modify this logic:

	for i in {0..7}

	do

		./worker $i 10 &

		if [ $? -ne 0 ] then

			echo "Erro ao iniciar o processo worker $i"

			exit 1

		fi

		sleep 1

	done

Thus, in the current logic, **the correct answer is 80.** 🚀