National Technical University of Athens (NTUA)
Τεχνολογίες διαχείρισης υπολογιστικών υποδομών και εφαρμογών σε υπηρεσιοστρεφείς αρχιτεκτονικές και περιβάλλοντα νεφών
Abstract
dc:descriptionThe evolution of Cloud Computing and the offerings of the Service Oriented Infrastructures (SOIs) in general, allowed the on-demand software and resource providing as a service. The effective usage and the efficient management of infrastructures are raising multiple and complex demands as far as the collection and the management of information generated from those infrastructures concerns. To this end, the monitoring techniques and mechanisms in SOIs and Cloud environments form an important component of the value chain of those infrastructures. This thesis is focused in this study of information management towards surpassing the limitations and restrictions that are introduced by the existence of different entities (Application Provider, Platform Provider and Infrastructure Provider) but also technical restrictions caused by this new architectural paradigm. More specifically, the analysis and design of innovative monitoring mechanisms is presented, managing information deriving from infrastructures as well as applications executed in those. Special focus has been given in the capability of re-configuration of the monitoring mechanism based on specific values of the collected information. This dynamic characteristic allows for a proactive violation detection of SLAs as well as the optimization of the infrastructure through better decision-making policies. The particular monitoring solution constitutes a service-oriented framework for data collection so much from the virtual and physical infrastructure. The policy of re-configuration is applied throughout this service framework which is evaluated with a real-life application scenario of a color correction video post-production. Moreover, a second innovative monitoring infrastructure for Cloud environments is presented which aims at collecting information from various sources and assess it in order for an optimization of all Cloud entities. Through this work we focus on the collection of energy consumption of the resources and the calculation of the eco-efficiency of the infrastructure. Based on power models, the definition of eco-efficiency is provided for all Cloud entities and forecasted values are being calculated. This modeling leads to the application of policies that improve the placement of virtual machines (VMs) on the Cloud infrastructure. The presented results are very encouraging with regard to the possibilities of management of resources taking into consideration the eco-efficiency of infrastructure. Finally, this thesis concludes with some additional contributions to the open source community, related with technologies and techniques of resource monitoring. In detail, an architecture and implementation of a monitoring plug-in of Nagios system is presented, which allows for a service-oriented operation of the pre-mentioned monitoring system. Furthermore, an experimental set-up of a Cloud-enabled elastic web hosting scenario is being described and evaluated, using only open-source software and applied in a private Cloud infrastructure.
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2012
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Κατσαρός, Γρηγόριος
- Katsaros, Gregory
Subjects
dc:subject × 14- Υπηρεσιοστρεφείς υποδομές
- Υπολογιστικό νέφος
- Επίβλεψη πληροφορίας
- Εικονικοποίηση
- Ενεργειακή απόδοση
- Service oriented infrastructures
- Cloud computing
- Information monitoring
- Virtualization
- Eco-efficiency
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/27131