raftleaderelection is a golang code - a raft election code which simulates the behaviour of raft election protocol. It ensures that only one leader will be present at a time in a cluster. It ensures one leader with consideration of all conditions that may occur.

#Usage

(note: In the examples below, $ is the shell prompt, and the output of the snippet follows "----------------"

$ go test
---------------------------------
PASS
ok    Raft   141.193s

This command executes test program which contains the test which checks the minory & majority failures. This program kills the servers one by one and checks the only one leader in maximum number of server killing bound.

$ go run main.go 0 11110 /home/rahul/IdeaProjects/cloud/src
--------------------------------
0 in follower
0 in candidate

The first parameter is id, second one is port and third parameter is the parent directory This program runs the program for the above parameters. The configuration for raft are written in configuration_.xml files and configuration for cluster package are written in serverlist_.xml Currently program is written for 5 servers which can be converted into more by modifying the configuration files for above two packages.

go get github.com/RahulWaniIITB/raftleaderelection

This package contains cluster layer which will be used in making & sending & receiving the messages.

This package contains raft leader election layer which will ensure that only one leader at a time. This part simulates the raft protocol.

raft_test.go contains the code for testing the software. This program by using raft package creates the raft instance and starts its execution. The raft instance created will eventually create the cluster instance which will bind network socket with each raft instance.

raftleaderelection is available free. So, anyone can use this repository or modify it.