Qazi, Farrukh (2020) Automating SLA enforcement in the cloud computing. PhD thesis, University of Warwick.

Preview

PDF WRAP_Theses_Qazi_2020.pdf - Submitted Version - Requires a PDF viewer. Download (4018Kb) | Preview

Request Changes to record.

Abstract

Cloud computing is playing an increasingly important role, not only by facilitating digital trading platforms but also by transforming conventional services from client-server models to cloud computing. This domain has given the global economic and technological benefits, it offers to both the service providers and service subscribers. Digital marketplaces are no longer limited only to trade tangible commodities but also facilitates enormous service virtualization across various industries. Software as a Service (SaaS) being the largest service segment, dominates the global cloud migration. Infrastructure as a Service (IaaS) and cloud-based application development also known as Platform as a Service (PaaS) are also next-generation computing platforms for their ultimate futuristic demand by both, public and private sector. These service segments are now hosted on cloud platforms to compute, store, and network, an enormous amount of service requests, which process data incredibly fast and economically.

Organizations also perform data analytics and other similar computing amenities to manage their business without maintaining on-premise computing infrastructures which are hard to maintain. This computing capability has extensively improved the popularity and increased the demand for cloud services to an extent, that businesses worldwide are heavily migrating their computing resources to these platforms. Diverse cloud service providers take the responsibility of provisioning such cloud-based services for subscribers. In return, a certain subscription fee is charged to them periodically and depending upon the service package, availability and security. On the flip side, such intensive technology shift and outsourcing reliance have also introduced scenarios that any failure on their part leads to serious consequences to the business community at large.

In recent years technology industry has observed critical and increased service outages at various cloud service providers(CSP) such as Amazon AWS, Microsoft, Google, which ultimately interrupts the entire supply chain and causes several well-known web services to be taken offline either due to a human error, failed change control implementation or in more recently due to targeted cyber-attacks like DDoS. These web-based solutions such as compute, storage, network or other similar services are provisioned to cloud service subscribers (CSS) platforms. Regardless of a cloud service deployment, a legal binding such as a Service Level Agreement (SLA) is signed between the CSP and CSS. The SLA holds a service scope and guarantees in case of failure. There are probabilities where these SLA may be violated, revoked, or dishonoured by either party, mostly the CSP.

An SLA violation along with an unsettled dispute leads to some financial losses for the service subscribers or perhaps cost them their business reputation. Eventually, the subscriber may request some form of compensation from the provider such as a service credit or a refund. In either case, the burden of proof lies with the subscribers, who have to capture and preserve those data or forensically sound system or service logs, supporting their claims. Most of the time, this is manually processed, which is both expensive and time-consuming. To address this problem, this research first analyses the gaps in existing arrangements. It then suggests automation of SLA enforcement within cloud environments and identifies the main properties of a solution to the problem covering various other avenues associated with the other operating environments. This research then subsequently proposes architectures, based on the concept of fair exchange, and shows that how intelligently the approach enforces cloud SLA using various techniques. Furthermore, by extending the research scope covering two key scenarios (a) when participants are loss averse and (b) when interacting participants can act maliciously. Our proposed architectures present robust schemes by enforcing the suggested solutions which are effective, efficient, and most importantly resilient to modern-day security and privacy challenges. The uniqueness of our research is that it does not only ensure the fairness aspect of digital trading but it also extends and logically implements a dual security layer throughout the service exchange. Using this approach protects business participants by securely automating the dispute resolutions in a more resilient fashion. It also shields their data privacy and security from diverse cyber challenges and other operational failures. These architectures are capable of imposing state-of-the-art defences through integrated secure modules along with full encryption schemes, mitigating security gaps previously not dealt with, based upon fair exchange protocols. The Protocol also accomplishes achieving service exchange scenarios either with or without dispute resolution. Finally, our proposed architectures are automated and interact with hardcoded procedures and verifications mechanism using a variant of trusted third parties and trusted authorities, which makes it difficult to cause potential disagreements and misbehaviours during a cloud-based service exchange by enforcing SLA.

Item Type:	Thesis or Dissertation (PhD)
Subjects:	H Social Sciences > HF Commerce Q Science > QA Mathematics > QA76 Electronic computers. Computer science. Computer software
Library of Congress Subject Headings (LCSH):	Cloud computing, Service-level agreements, Data encryption (Computer science), Online stockbrokers, Web services, Electronic commerce
Official Date:	December 2020
Dates:	Date Event December 2020 UNSPECIFIED
Institution:	University of Warwick
Theses Department:	Department of Computer Science
Thesis Type:	PhD
Publication Status:	Unpublished
Supervisor(s)/Advisor:	Jhumka, Arshad, 1974-
Sponsors:	Engineering and Physical Sciences Research Council
Format of File:	pdf
Extent:	xii, 136 leaves : illustrations
Language:	eng

Request changes or add full text files to a record

Repository staff actions (login required)

View Item

Downloads