CloudSim vs CloudSim Plus: Dynamic Workload Simulation & Performance Analysis
This project demonstrates a performance comparison between CloudSim and CloudSim Plus using dynamically generated cloud simulation scenarios.
The system continuously generates random workloads by varying parameters such as:
Host CPU cores
Virtual machine (VM) cores
Number of VMs
Number of cloudlets (tasks)
Workload length
Each scenario is executed in both CloudSim and CloudSim Plus environments. The simulation process includes:
1. Creating a datacenter with available resources
2. Allocating virtual machines (VMs)
3. Assigning cloudlets (tasks) to VMs
4. Running the simulation engine
5. Collecting performance metrics
The following metrics are analyzed:
Execution Time
Average CPU Time
Makespan (total completion time)
Throughput (tasks completed per unit time)
CloudSim Plus additionally provides structured output and a side-by-side comparison of results, making it easier to evaluate performance differences.
Key Observations:
CloudSim Plus generally shows lower execution time
Reduced makespan compared to CloudSim
Improved throughput in most scenarios
Cleaner and more structured output format
This project highlights how dynamic workload simulation can be used to evaluate cloud performance across multiple scenarios.
Conclusion:
CloudSim Plus provides better performance, scalability, and usability compared to the original CloudSim framework, making it more suitable for modern cloud computing research.
Tools & Technologies Used:
Java
CloudSim (v3.x)
CloudSim Plus
Maven
WSL (Linux environment)