Cloud Computing: Artificial Intelligence for IT Operations (AIOps) (IN2107)

Prof. Dr. Michael Gerndt, Anshul Jindal, Mohak Chadha, Paolo Notaro, Vladimir Podolskiy

Motivation

Large-scale distributed systems for applications such as the cloud are prone to failures due to the occurrence of faults in the software and the hardware. As a result, cloud service providers employ site reliability and DevOps engineers to ensure high availability of services. However, due to the increasing complexity of the underlying systems manual operation has become infeasible. To this end, cloud system operators employ artificial intelligence-based tools and techniques (i.e. AIOps) for system management tasks. AIOps is an emerging interdisciplinary domain arising in the intersection between the research areas of machine learning, big data, streaming analytics, and the management of IT operations. 

This year's seminar will focus on the usage of ML/DL techniques for smart cloud operations. The goal of this seminar is to investigate the usage of these techniques for the management of different aspects of cloud systems.

The topics that will be covered are: 

• Online Failure Prediction
• Root-cause Analysis and Diagnosis
• Fault Detection/Anomaly Detection
• Power-efficient Datacenter Management
• Nonconventional virtualization for clouds - technologies and use-cases
• Hybrid resource management in cloud systems
• Forecasting resource scheduling/provisioning
• Fault Tolerance and resiliency in cloud computing
• Privacy-preserving Real-time Anomaly Detection
• Predicting Node Failures in an Ultra-Large-Scale Cloud Computing
• Advanced Resource provisioning in cloud
• Blockchain and Cloud Computing

The students will select material for the presentation, provide a report, and give a presentation of 30 minutes. The management of the seminar is done through Moodle.