In recent years the evolution and the widespread adoption of virtualization, service-oriented architectures, autonomic and utility computing have converged letting a new paradigm to emerge: The Cloud Computing. Cloud Computing aims at streamlining the on-demand provisioning of software, hardware, and data as services, providing end-user with flexible and scalable services accessible through the Internet. Since the Cloud offer is currently becoming wider and more attractive to business owners, the development of efficient resource provisioning policies for Cloud-based services becomes increasingly challenging. Indeed, modern Cloud services operate in an open and dynamic world characterized by continuous changes where strategic interaction among different economic agents takes place. This thesis aims to study the hourly basis service provisioning and capacity allocation problem through the formulation of a mathematical model based on noncooperative-game-theoretic approach. We take the perspective of Software as a Service (SaaS) providers which want to minimize the costs associated with the virtual machine instances allocated in a multi-IaaSs (Infrastructure as a Service) scenario, while avoiding incurring in penalties for requests execution failures and providing quality of service guarantees. SaaS providers compete and bid for the use of infrastructural resources, while the IaaSs want to maximize their revenues obtained providing virtualized resources. The problem has been modeled as a Generalized Nash Equilibrium Problem (GNEP). Thanks to a deep analysis of the game under study, we demonstrate the social equilibria existence for the corresponding generalized potential game. The best-reply solution is pursued heuristically with the implementation of two different algorithms suitable for a distributed implementation. We demonstrate the effectiveness of our approach by performing numerical analyses, considering multiple workloads and system configurations. Results show that our algorithms are scalable. Equilibria efficiency is quantitatively analysed in terms of Price of Anarchy (PoA) and Individual Worst Case (IWC) for both SaaSs and IaaSs, indicating the advantage of solving the equilibrium problem with the second approach (improvements by up tp 40%). Furthermore, a multiple IaaS analysis points out the SaaS benefits in exploiting multiple IaaS deployment of applications and redistribution of traffic simultaneously: in terms of efficiency, performance improves by 40%, while, in terms of costs, SaaSs average saving ranges from 10% up to 40% compared to single IaaS architectures.
Negli ultimi anni l'evoluzione e la diffusa adozione di virtualizzazione, architetture orientate ai servizi, autonomic e utility computing sono confluiti in un nuovo paradigma emergente: il Cloud Computing. Il Cloud Computing mira a semplificare la fornitura on-demand di software e hardware, fornendo agli utenti finali servizi flessibili e scalabili accessibili tramite Internet. Poichè ultimamente l'offerta Cloud si sta ampliando, diventando più attrattiva per le aziende, lo sviluppo di politiche efficienti per la distribuzione delle risorse per servizi basati sul Cloud diventa sempre più complesso. Infatti, le moderne infrastrutture Cloud operano in un mondo dinamico e caratterizzato da continui cambiamenti e interazioni tra diversi concorrenti. L'obiettivo di questa tesi è lo studio del problema della fornitura oraria di servizi e l'allocazione di risorse attraverso la formulazione di un modello matematico basato sui concetti della teoria dei giochi non cooperativi. Viene considerata la prospettiva dei fornitori di Sofware as a Service (SaaS) che vogliono minimizzare i costi associati alle macchine virtuali allocate in uno scenario multi-IaaS (Infrastructure as a Service), evitando sanzioni causate dalla mancata esecuzione delle richieste e garantendo la qualità del servizio. I SaaS competono fra loro facendo offerte per l'utilizzo delle infrastrutture, mentre gli IaaS vogliono massimizzare i propri introiti ottenuti fornendo risorse virtualizzate. Il modello è costituito da un problema di equilibrio di Nash generalizzato. Grazie ad un'accurata analisi del problema studiato, dimostriamo l'esistenza di equilibri sociali per il corrispondente gioco generalizzato a potenziale. La soluzione del best-reply è ottenuta euristicamente con l'implementazione di due differenti algoritmi distribuiti. L'efficacia del nostro approccio è stata dimostrata compiendo test numerici, considerando differenti configurazioni di carico e di sistema. I risultati mostrano che gli algoritmi sono scalabili. L'efficienza degli equilibri è quantitativamente analizzata in termini di Price of Anarchy (PoA) e Individual Worst Case (IWC) di SaaS e IaaS, indicando il vantaggio nella risoluzione del problema con il secondo approccio (ottenendo miglioramenti fino al 40%). Inoltre, l'analisi multi-IaaS evidenzia i benefici dei SaaS nello sfruttare il dislocamento delle applicazioni e la distribuzione del traffico su molteplici IaaS: in termini di efficienza, le performance migliorano del 40%, in termini di costi, il risparmio medio per i SaaS oscilla tra il 10% e il 40% rispetto al caso mono IaaS.
Service provisioning problem in cloud and multicloud systems : a generalized Nash equilibrium model
SAVI, ANNA
2011/2012
Abstract
In recent years the evolution and the widespread adoption of virtualization, service-oriented architectures, autonomic and utility computing have converged letting a new paradigm to emerge: The Cloud Computing. Cloud Computing aims at streamlining the on-demand provisioning of software, hardware, and data as services, providing end-user with flexible and scalable services accessible through the Internet. Since the Cloud offer is currently becoming wider and more attractive to business owners, the development of efficient resource provisioning policies for Cloud-based services becomes increasingly challenging. Indeed, modern Cloud services operate in an open and dynamic world characterized by continuous changes where strategic interaction among different economic agents takes place. This thesis aims to study the hourly basis service provisioning and capacity allocation problem through the formulation of a mathematical model based on noncooperative-game-theoretic approach. We take the perspective of Software as a Service (SaaS) providers which want to minimize the costs associated with the virtual machine instances allocated in a multi-IaaSs (Infrastructure as a Service) scenario, while avoiding incurring in penalties for requests execution failures and providing quality of service guarantees. SaaS providers compete and bid for the use of infrastructural resources, while the IaaSs want to maximize their revenues obtained providing virtualized resources. The problem has been modeled as a Generalized Nash Equilibrium Problem (GNEP). Thanks to a deep analysis of the game under study, we demonstrate the social equilibria existence for the corresponding generalized potential game. The best-reply solution is pursued heuristically with the implementation of two different algorithms suitable for a distributed implementation. We demonstrate the effectiveness of our approach by performing numerical analyses, considering multiple workloads and system configurations. Results show that our algorithms are scalable. Equilibria efficiency is quantitatively analysed in terms of Price of Anarchy (PoA) and Individual Worst Case (IWC) for both SaaSs and IaaSs, indicating the advantage of solving the equilibrium problem with the second approach (improvements by up tp 40%). Furthermore, a multiple IaaS analysis points out the SaaS benefits in exploiting multiple IaaS deployment of applications and redistribution of traffic simultaneously: in terms of efficiency, performance improves by 40%, while, in terms of costs, SaaSs average saving ranges from 10% up to 40% compared to single IaaS architectures.| File | Dimensione | Formato | |
|---|---|---|---|
|
2013_04_Savi.pdf
solo utenti autorizzati dal 03/04/2016
Descrizione: Testo della tesi
Dimensione
2.14 MB
Formato
Adobe PDF
|
2.14 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/77641