PROJECT TITLE :

Model-Based Thermal Anomaly Detection in Cloud Datacenters Using Thermal Imaging - 2018

ABSTRACT:

The growing importance, giant scale, and high server density of high-performance computing datacenters build them vulnerable to attacks, misconfigurations, and failures (of the cooling also of the computing infrastructure). Such surprising events typically lead to thermal anomalies - hotspots, fugues, and coldspots - that impact the cost of operation of datacenters. A model-primarily based thermal anomaly detection mechanism, which compares expected (obtained using heat-generation and -extraction models) and observedthermal maps (obtained using thermal cameras) of datacenters, is proposed. Additionally, a novel Thermal Anomaly-aware Resource Allocation (TARA) is intended to induce a time-varying thermal fingerprint (thermal map) of the datacenter therefore to maximize the detection accuracy of the anomalies. As shown via experiments on a small-scale testbed as well as via trace-driven simulations, such model-based thermal anomaly detection answer along with TARA significantly improves the detection likelihood compared to anomaly detection when scheduling algorithms such as random, spherical robin, and best-fit-decreasing are utilized.