Load Balancing Script Usage Guide

Overview

This Python script is part of a load balancing problem, which is a common issue in distributed computing where work is distributed among multiple computers or servers to achieve optimal resource utilization, minimize response time, and avoid overload. Specifically, this script is part of a project that aims to find the optimal schedule for cloud load balancing to keep a service running in the cloud at reasonable cost by reducing the expense of running cloud servers while minimizing risk and human time due to migrating and doing balance sleeping models across servers.

The objective function:

• number_of_active_servers: Minimize the total cost of running servers
• number_of_migrations: Minimize the total cost of migrating users from one server to another to reduce the risk and human time
• max_sleeping_workload: Minimizing the upper bound of the number of sleeping process on each server to balance the sleeping process across servers

p.s.: These three objectives are combined into one objective function by order of importance in distributed computing.

The constraints:

• The number of processes running on each server must less than max_process_per_server
• Each user is assigned to exactly one server and this server must be active
• The number of processes sleeping on each server must less than max_sleeping_workload

The decision variables:

• assign_user_to_server_vars: assign_user_to_server[i,j] = 1 if user i is assigned to server j, 0 otherwise
• active_var_by_server: active_servers[j] = 1 if server j is active, 0 otherwise

Requirements

• Python 3.10 or later
• docplex module 2.25.236 or later You can install the required Python module using pip or conda, you should notice that by this method, the module will be community version.

  conda install docplex==2.25.236

How to Use

1. Ensure you have the necessary Python version and modules installed.
2. Run the script using Python:

   python load_balancing.py

The script will print the solution to the console and also save it as a JSON file named solution.json.

Customization

The script uses predefined data for the servers and users. If you want to use your own data, you can modify the SERVERS and USERS variables in the script. The SERVERS variable is a list of server names. The USERS variable is a list of tuples, where each tuple represents a user and contains the user's ID, the number of running processes, the number of sleeping processes, and the current server.

For example:

SERVERS = ["server1", "server2", "server3"]

USERS = [("user1", 2, 1, "server1"),
         ("user2", 0, 2, "server1"),
         ("user3", 0, 4, "server2")]

This represents a scenario with three servers and three users. User1 is currently on server1 and has 2 running processes and 1 sleeping process. User2 is also on server1 but has no running processes and 2 sleeping processes. User3 is on server2 and has no running processes and 4 sleeping processes.