Distributed chat based on Chang-Roberts algorithm
This program is a semestral project on MI-PYT and MI-DSV courses. It is implemented in Python 3.7 and uses AsyncIO.
Since a few new AsyncIO features were used in this program, it is necessary to have Python 3.7 or higher to run it.
This application is a CLI chat room program. However, it does not use a simple server-client way. All nodes are clients, any node can be a server from this point of view.
For better understanding, the server (node that manages the user messages) will be referred to as the leader and the AsyncIO component of the application that listens on the given port will be refered to as the server further on in the docuentation. Meaning that any node has a server, however only one node in the chatroom is the leader.
The chatroom is a network ring (a node only knows its next and previous node), however all nodes know the leader node as well in order to send it a message. The first node that creates the chatroom is automtically its leader. When the leader dies or logs out, the new leader is elected using Chang-Roberts leader election algorithm.
When the user types a message, the node sends it to the leader node and the leader distributes it to all the participants (the leader knows them all).
When a node wants to join the chat room, all it needs is one address of any node that is already in the room.
When a node logs out or dies without informing the others, its next and previous nodes detect it and repair the ring.
The program is uploaded on PyPI.
For the newest version, just type:
python -m pip install soucevi1_dist_chat
For any specific version, type:
python -m pip install soucevi1_dist_chat==<version>
You can also clone and use the GitHub repository.
Creating the chatroom
To run the application as the leader node after the installation, type:
soucevi1_dist_chat -l -a <IP> -p <port> -n <your name>
Where <IP> and <port> is the IP address and port the node will listen on, <your name> is the username of your choice and -l stands for being the leader.
Joining the existing chatroom
To join the existing chatroom, you need to know IP <nIP> and port <nPort> of any node that is already in the chatroom:
soucevi1_dist_chat -a <IP> -p <port> -n <your name> -na <nIP> -np <nPort>
In both cases you can specify the option -v that stands for verbose. Without -v, the application logs into file (chat_<port>.log). With -v the application logs to the file as well as to the stdout.
The format of the log message is:
Where TS is Lamport logical clock timestamp.
Once you start the leader, you won’t be able to do anything until the first connection of another node is made. After that, you get a prompt that looks like this:
Where you can write your message and press Enter to send it. Received message looks simillar:
> Name(IP:port)[Lamport timestamp]: Hello from Name
To exit the application, instead of your message simply write //exit.
For help, run:
You can take a look at the documentation on this website.
Or, you can build it using Sphinx:
$ cd docs $ make html
You can replace html in the command with whatever Sphinx currently supports. Beware that some third party Sphinx extensions might be required (listed in docs/requirements.txt).
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.