Outage-Watch: Early Prediction of Outages using Extreme Event Regularizer

Foundations of Software Engineering (ESEC/FSE)

Publication date: December 6, 2023

Shubham Agarwal, Sarthak Chakraborty, Shaddy Garg, Sumit Bisht, Chahat Jain, Ashritha Gonuguntla, Shiv Saini

Cloud services are omnipresent and critical cloud service failure is a fact of life. In order to retain customers and prevent revenue loss, it is important to provide high reliability guarantees for these services. One way to do this is by predicting outages in advance, which can help in reducing the severity as well as time to recovery. It is difficult to forecast critical failures due to rarity of these events. Moreover, critical failures are ill-defined in terms of observable data. Our proposed method, Outage-Watch, defines outages as deteriorations in the Quality of Service (QoS) captured by a set of metrics. This definition allow us to capture a large set of critical service outages. Outage-Watch detects outages in advance by using current system state to predict whether the QoS metrics will cross a threshold. A mixture of gaussians is used to model the distribution of the QoS metrics for flexibility and an extreme event regularizer helps in improving learning in tail of the distribution. An outage is predicted if the probability of any one of the QoS metrics crossing threshold changes significantly. Our evaluation on a real-world SAAS company dataset shows that Outage-Watch significantly outperforms traditional methods with an average AUC of 0.98. Additionally, Outage-Watch detects all the outages exhibiting a change in service metrics and reduces the Mean Time To Detection (MTTD) of outages by up to 88% when deployed in an enterprise cloud-service system, demonstrating efficacy of our proposed method.

Learn More