PREDICTIVE MAINTENANCE IN CLOUD COMPUTING AND DEVOPS: ML MODELS FOR ANTICIPATING AND PREVENTING SYSTEM FAILURES

Abstract

This paper aims to review the current trends and use of machine learning (ML) models in the predictive maintenance of IT infrastructure, especially cloud computing and DevOps ecosystems. The objective is to foresee systemic risks and guard against their occurrence by reviewing operational data and establishing succeeding patterns of poor system performance. Experiments are performed with accurate logs of system and service activities from a cloud service provider. At the same time, the developed ML models were established to provide reasonably accurate estimates of failures, hence minimizing the possibility of 'random' failures. In real-life events, the model proved useful in anticipating future system loads and interruptions that would injure the system. Some of the addressed challenges include the real-time processing of data streams and the scalability of data processing. The paper results show that ML models can underpin the increasingly dependable system, optimize operations in a cloud-computing environment and DevOps, and propose an innovative strategy for system preservation and failure anticipation

Published Date: 12-11-2023