Skip to content

Implementations for Pastry Protocol and Chord Distributed Hash Tables in Python

License

Notifications You must be signed in to change notification settings

DivyanshuSaxena/Distributed-Hash-Tables

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

25 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Distributed Hash Tables

Implementation of the Pastry (paper) and Chord (paper) DHT Protocols, as part of assignment for course COL819 - Spring 2020 at IIT Delhi.

Repository Structure

.  
├── chord_node.py  
├── chord.py  
├── LICENSE  
├── links.dat  
├── modules  
│   ├── __ init__.py  
│   └── network.py  
├── pastry_node.py  
├── pastry.py  
└── README.md  

For further details, check the docstring in each of the python files.

How to run

Pastry Object Location Service

>>> python pastry.py <num-nodes-in-network> <whether-to-read-network-configuration-from-file (0/1)>

Parameters

The default parameters are as follows:

l = 6  # Length of hash code
b = 4  # Parameter to set Leaf Set and Neighborhood Set
num_points = 10000  # Number of data points in Pastry
num_queries = 1000000  # Number of queries made

To change the parameters, go to pastry.py

Chord Peer to Peer DHT

>>> python chord.py <num-nodes-in-network> <whether-to-read-network-configuration-from-file (0/1)>

Parameters

The default parameters are as follows:

l = 6  # Hashing scheme for Nodes (Not required for Chord DHT, but to verify correctness)
m = 24  # Number of entries in Finger Table
num_points = 10000   # Number of data points to store in Chord
num_queries = 1000000  # Number of queries made on the DHT

To change the parameters, go to chord.py

Network Simulation

Both of the services, Pastry and Chord are implemented, using an underlying network simulation, with a definite measure of geographical distance (or proximity metric) between the nodes. The Network has been simulated by keeping a graph of interconnected vertices. Each vertex may logically correspond to a Pastry/Chord Node. The physical distance between two Pastry/Chord Nodes is hence kept as the distance between the corresponding vertices in the network graph.

At each run, a new Network is initiated and the links (or edges) are stored in a file links.dat. The network configuration can be read from this file in a subsequent run, by specifying the respective argument. See the section How to run.

The Network class implemention and code can be found in the modules.network script. The Network Class utilizes the Node class definition, which is the base class for the PastryNode and ChordNode classes, used in Pastry and Chord respectively.

The Network instance has been used for three major functions:

def get_node(self, node_id):
    """Get the node at node id on the nodes array

    Arguments:
        node_id {Integer} -- Hash of the node

    Returns:
        Node -- Associated Node
    """
    return self.nodes[node_id]

def is_alive(self, n):
    """Checks if a given node n is alive in the network

    Arguments:
        n {Integer} -- Node Id of the node

    Returns:
        Boolean -- True if node is alive else False
    """
    if n in self.nodes:
        return True
    return False

def proximity(self, n1, n2):
    """Define the proximity metric between two Node instances on the network

    Currently using the modulo additive inverse for the metric

    Arguments:
        n1 {Integer} -- Hash of node n1
        n2 {Integer} -- Hash of node n2

    Returns:
        Integer -- the proximity metric (Returns -1 if node not alive)
    """
    try:
        # ... Code to get the distance between the associated logical vertices in the graph
        return distance
    except:
        return -1

Contributing

Feel free to fork, make your changes and submit a pull request on this repo.

License

This project is licensed under the MIT License - see the LICENSE.md file for details

About

Implementations for Pastry Protocol and Chord Distributed Hash Tables in Python

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages