Navigation Menu

Skip to content
This repository has been archived by the owner on Feb 27, 2023. It is now read-only.

KuangjuX-Archived/TrivalTCPStack

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TrivialTCP

TrivialTCP is an open source TCP library in user mode implmented in pure C.

We use UDP to simulate the IP layer of the kernel to send and receive packets. TrivialTCP is the core part of this project, which can send and receive reliably.

Otherwise, we have also implemented high-performance components such as thread pools, epolls, and coroutines in user mode.

At the top layer, we expose APIs similar to the POSIX standard to users, so that we can provide reliable transmission services for applications.

Requirements

  • GCC
  • Dokcer

Start

In project directory, we can execute following command:

docker build -t tcp .

So we have a docker image named tcp.

As follow, we have to create my own network configuration:

docker network create --subnet=10.0.0.0/16 mynetwork

Before create the server/client containers, you should know the absolute path of this project in your computer. You can use the pwd command to get the path.

And then we can run our container by static ip address and bind this project in a double-way between your host and containers:

NOTICE! Replace {project path on your host} by your path:

sudo docker run --cap-add NET_ADMIN -itd -p 22 --name server --net mynetwork --ip 10.0.0.3 --hostname server -v {project path on your host}:/tjutcp tcp /bin/bash
sudo docker run --cap-add NET_ADMIN -itd -p 22 --name client --net mynetwork --ip 10.0.0.2 --hostname client -v {project path on your host}:/tjutcp tcp /bin/bash

The following commands could attach you in to each container:

 sudo docker exec -it client  /bin/bash
 sudo docker exec -it server  /bin/bash

execute make server in server container and execute make client in client container under the /tjutcp folder, we can run the project easily.

Workflows:

  • Environment
  • Thread Pool
  • Channel
  • Coroutine
  • Epoll
  • Connection Management
  • Timer
  • System API: bind, listen, accept, connect, recv, send, close
  • Reliable Transmission
  • Flow Control
  • Congestion Control

Differnes

We use 10.0.0.3 as server's ip address instead of 10.0.0.1 because this address is reported occupied.

Other

  • Once create the containers by the params -itd, they keep running in the background.
  • Get more information on the Docker website. Destroy all running containers:
sudo docker container prune
  • Use valgrind to check the status of memory leak:
valgrind --tool=memcheck (--leak-check=full) {exectuable file}

References