کنفرانس مهندسی برق ایران

صفحه اصلی / سی و دومین کنفرانس بین المللی مهندسی برق

Cloudy: A Pythonic Cloud Simulator

نویسندگان :

Ahmad Siavashi¹ Mahmoud Momtazpour²

1- دانشگاه صنعتی امیرکبیر (پلی تکنیک تهران) 2- دانشگاه صنعتی امیرکبیر (پلی تکنیک تهران)

کلمات کلیدی :

cloud computing،simulation،modeling،python

چکیده :

Cloud computing is vital for modern services, but optimizing cloud environments can be challenging, necessitating simulation tools. Existing cloud simulators, often written in Java, struggle to adapt to evolving research needs and integrate with modern libraries, especially in AI and LP. To address these challenges, we present Cloudy, a Python-based simulator for modeling modern cloud environments. Python's simplicity, readability, and extensive library support make Cloudy an accessible yet powerful tool for comprehensive cloud simulations. It leverages Python's seamless integration with AI and LP libraries for advanced experimentation and supports modern cloud peripherals, such as GPUs. Moreover, Cloudy accommodates various cloud service models, making it highly versatile for different cloud-based scenarios. Compared to the established CloudSim simulator, Cloudy demonstrated a significant reduction in memory usage and simulation time by 15.6x and 2.75x, respectively.