PROJECT TITLE :

Praxi: Cloud Software Discovery That Learns From Practice

ABSTRACT:

Users of cloud systems have the responsibility of monitoring the software that is running on their containers and virtual machines (VMs) in order to ensure compliance, security, and efficiency in light of the rapidly evolving cloud landscape of today, which embraces continuous integration and delivery. Traditional approaches to this problem rely on rules that are manually created to identify software installations and modifications. These rules, however, require knowledgeable authors and are frequently impossible to maintain. In recent years, automated methods for the discovery of software have come into existence. In some approaches, examples of software are used to teach machine learning models so that they can determine which software is already present on a computer system. Others utilize the knowledge of packaging practices to assist in discovery without requiring any prior training; however, these practice-based methods are unable to provide information that is precise enough to perform discovery on their own. This article introduces Praxi, a new software discovery method that builds upon the strengths of prior approaches by combining the accuracy of learning-based methods with the efficiency of practice-based methods. Specifically, Praxi combines the accuracy of learning-based methods with the efficiency of practice-based methods. Praxi is able to correctly classify installations at least 97.6 percent of the time in tests using samples collected on real-world cloud systems, while also running 14.8 times faster and using 87 percent less disk space than a similar learning-based method. These results were obtained from tests. This article makes a quantitative comparison of Praxi to systematic rule-, learning-, and practice-based methods by making use of a diverse software dataset. It then discusses the best applications for each methodology.