Computing and Information Systems - Theses

Permanent URI for this collection

Search Results

Now showing 1 - 2 of 2
  • Item
    Thumbnail Image
    Anomaly-aware Management of Cloud Computing Resources
    Kardani Moghaddam, Sara ( 2019)
    Cloud computing supports on-demand provisioning of resources in a virtualized, shared environment. Although virtualization and elasticity characteristics of cloud resources make this paradigm feasible, however, without efficient management of resources, the cloud system’s performance can degrade substantially. Efficient management of resources is required due to the inherent dynamics of cloud environment such as workload changes or hardware and software functionality such as hardware failures and software bugs. In order to meet the performance expectations of users, a comprehensive understanding of the performance dynamics and proper management actions is required. With the advent of data analysis techniques, this goal can be achieved by analyzing large volumes of monitored data for discovering abnormalities in the performance data. This thesis focuses on the anomaly aware resource scaling mechanisms which utilize anomaly detection techniques and resource scaling mechanism in the cloud to improve the performance of the system in terms of the quality of service and utilization of resources. It demonstrates how anomaly detection techniques can help to identify abnormalities in the behaviour of the system and trigger relevant resource reconfiguration actions to reduce the performance degradations in the application. The thesis advances the state-of-the-art in this field by making following contributions: 1. A taxonomy and comprehensive survey on performance analysis frameworks in the context of cloud resource management. 2. An Isolation-based anomaly detection module to identify performance anomalies in web based applications considering cloud dynamics. 3. An Isolation based iterative feature refinement to remove unrelated and noisy features to reduce the complexity of anomaly detection process in high-dimensional data. 4. A joint anomaly aware resource scaling mechanism for cloud hosted application. The approach tries to identify both the anomaly event and the root cause of the problem and trigger proper vertical and horizontal scaling actions to avoid or reduce performance degradations. 5. An adaptive Deep Reinforcement Learning (DRL) based scaling framework which leverages the knowledge of anomaly detection module to decide on proper decision making epochs. The scaling actions are encoded in DRL action space and the knowledge of actions values are obtained by training multi-layer Neural Networks.
  • Item
    Thumbnail Image
    A big data infrastructure for real-time traffic analytics on the cloud
    Gong, Yikai ( 2019)
    With the increasing urbanisation occurring globally, cities are facing unprecedented challenges. One major challenge is related to traffic and the increasingly common congestion issues that arise in cities. At the same time, digital data is being created across all walks of life by industry, governments and society more generally. The term "big data'' has now entered common vernacular. Big data can include officially captured data, e.g. from traffic measurement systems from government organisations such as VicRoads in Australia, as well as other forms of data generated by the population at large, e.g. social media. This thesis explores the unique characteristics of traffic related data and focuses on the development and evaluation of an underpinning Cloud-based platform that can tackle some of the unique big data challenges related to such data. In particular, the thesis focuses on challenges related to the volume, velocity and variety of traffic data. We explore how different forms of data including official sensor data such as the Sydney Coordinated Adaptive Traffic System (SCATS) that is widely rolled out across Victoria and supported by VicRoads can be processed in real time, as well as how social media data such as Twitter can be used as a cheaper proxy for SCATS to better understand traffic in cities. We also develop novel real-time clustering algorithms that tackle the unique spatial and temporal aspects of traffic related data.