Abstract
Cloud Computing has established itself as an efficient and cost-effective paradigm for the execution of web-based applications and scientific workloads that need elasticity and on-demand scalability capabilities. However, the evaluation of novel resource provisioning and management techniques is a major challenge due to the complexity of large-scale data centers. Therefore, Cloud simulators are an essential tool for academic and industrial researchers to investigate the effectiveness of novel algorithms and mechanisms in large-scale scenarios. This paper proposes CloudSim 7G, the seventh generation of CloudSim, which features a re-engineered and generalized internal architecture to facilitate the integration of multiple CloudSim extensions within the same simulated environment. As part of the new design, we introduced a set of standardized interfaces to abstract common functionalities and carried out extensive refactoring and refinement of the codebase. The result is a substantial reduction in lines of code with no loss in functionality, significant improvements in run-time performance and memory efficiency (up to 25% less heap memory allocated), as well as increased flexibility, ease-of-use, and extensibility of the framework. These improvements benefit not only CloudSim developers but also researchers and practitioners using the framework for modeling and simulating next-generation Cloud Computing environments.
Introduction
Since its debut in 2008, CloudSim has become the de-facto standard for simulating cloud data centers. However, as cloud computing evolved to include Software-Defined Networking (SDN), containers, and serverless models, the ecosystem became fragmented. Researchers created "single-feature" extensions (modules) that were often incompatible with one another.
As shown in the timeline above, the proliferation of concepts—from Virtual Machines to Quantum-as-a-service—created a significant "architectural debt." For example, simulating a container-based application running on a power-aware virtual machine within a specific network topology often required "hacking" disparate modules together, leading to massive code duplication and performance bottlenecks. CloudSim 7G addresses this by providing a unified base layer where these technologies can coexist and interact seamlessly.
Architectural Refactoring and Methods
The development of CloudSim 7G involved a massive refactoring of the existing codebase, removing over 13,000 lines of redundant code. Key technical changes include:
- Standardized Interfaces: The introduction of
HostEntityandGuestEntityinterfaces abstracts the behavior of physical machines and virtual components. This allows the simulator to treat VMs and containers through a common logic, enabling native support for nested virtualization (e.g., containers inside VMs). - Unified Selection Policies: Placement and migration logic, previously scattered across different packages, were consolidated into a single
SelectionPolicyinterface. This reduced 26 redundant classes down to just 11. - Engine Optimization: The core simulation engine was modernized by replacing inefficient custom linked lists (with $O(n)$ time complexity for event dispatching) with Java’s
PriorityQueue, achieving $O(\log n)$ complexity. - Virtualization Overhead Modeling: A new parameter was introduced to model the real-world performance "tax" that occurs when data traverses virtual interfaces, a crucial detail for accurate nested virtualization studies.
Performance Evaluation
To validate the performance gains, the authors evaluated CloudSim 7G using real-world CPU utilization data from PlanetLab Workload Traces and complex task dependencies modeled via Scientific Workflows (DAX format).
The benchmarks comparing CloudSim 6G to 7G revealed significant efficiencies:
- Memory Efficiency: CloudSim 7G utilizes up to 25% less total heap memory in large-scale simulations.
- Execution Speed: Run-time performance improved by 5% to 12% across various standard power-aware algorithms.
- Integration Capability: The authors successfully demonstrated a hybrid case study involving networked cloudlets, containers, and virtualization overhead—a scenario that was previously nearly impossible to simulate in a single environment.
Discussion and Key Insights
By removing over 13,000 lines of code, the framework's flexibility actually increased. For instance, the ContainerCloudSim module saw a 64% reduction in lines of code through deduplication. Furthermore, moving to modern Java standards (JDK 21) and replacing synchronized data structures—which were unnecessary for the single-threaded core engine—provided an "out-of-the-box" performance boost for researchers upgrading from older versions.
Conclusion and Future Work
While CloudSim 7G represents a significant leap forward in cloud simulation technology, several areas for future improvement remain. The authors identify the need for more realistic multi-core CPU scheduling models and the introduction of dedicated interfaces for AI/ML accelerators to support modern resource orchestration research. Additionally, moving from MIPS-based performance metrics toward clock frequency-based modeling is suggested to achieve higher simulation fidelity.
Source: Andreoli, R., et al. "CloudSim 7G: An Integrated Toolkit for Modeling and Simulation of Future Generation Cloud Computing Environments." arXiv:2408.13386v4, 2025.
Summary updated