Aristotle University Of Thessaloniki (AUTH)
Μοντελοποίηση, προσομοίωση και ανάλυση της απόδοσης και του κόστους υπολογιστικών νεφών
Abstract
dc:descriptionCloud Computing, just a niche market for early adopters with adventurous temperaments a few years ago, has evolved into a multi-billion dollar business and has established itself as the major computing platform for all kinds of data analysis in the exa scale. Ubiquitous access, to a virtually infinite pool of computing resources has spurred a revolution in online services with more and more old-school data-analytics “players” moving to the cloud and seeking a piece of the cloud computing pie. As the clouds grow, so do the scores of clients looking to move their businesses to the cloud trying to incorporate greater levels of flexibility, reliability and performance into their infrastructure. A need therefore arises, the need for better and more flexible scheduling of the cloud's resources, which are fluid and dynamic in nature. These resources, have levels of heterogeneity that are unheard of in classical distributed systems. In this thesis, our aim is, to explore how current scheduling methods can be adapted and optimized for use in the clouds and also develop new, modern methods that are better suited for cloud environments. To that end, we have partitioned this thesis into two sections, in the first we explore the use of gang scheduling methodologies while in the second we introduce novel scheduling schemes for meta-heuristic optimization in bag-of-tasks (BoT) scheduling. The contribution of this thesis lies in the adaptation and exploration of multiple methodologies, both well established and novel ones, in the scheduling of different families of parallel applications in dynamic single, multi, and inter-cloud computing models with varying degrees of performance, cost and size heterogeneity.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2015
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Moschakis, Ioannis
- Μοσχάκης, Ιωάννης
Subjects
dc:subject × 16- Υπολογιστικά νέφη
- Εικονικές μηχανές
- Χρονοδρομολόγηση παράλληλων εργασιών
- Μετα-ευρετικοί αλγόριθμοι
- Cloud computing
- Virtual machines
- Parallel job scheduling
- Meta-heuristic algorithms
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Natural Sciences
- Computer and Information Sciences
- Engineering and Technology
- Electrical Engineering, Electronic Engineering, Information Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/36317