Current Issue : October - December Volume : 2014 Issue Number : 4 Articles : 4 Articles
Cloud computing plays an increasingly important role in realizing scientific applications by offering virtualized\ncompute and storage infrastructures that can scale on demand. This paper presents a self-configuring adaptive\nframework optimizing resource utilization for scientific applications on top of Cloud technologies. The proposed\napproach relies on the concept of utility, i.e., measuring the usefulness, and leverages the well-established principle\nfrom autonomic computing, namely the MAPE-K loop, in order to adaptively configure scientific applications. Therein,\nthe process of maximizing the utility of specific configurations takes into account the Cloud stack: the application layer,\nthe execution environment layer, and the resource layer, which is supported by the defined Cloud stack configuration\nmodel. The proposed framework self-configures the layers by evaluating monitored resources, analyzing their state,\nand generating an execution plan on a per job basis. Evaluating configurations is based on historical data and a utility\nfunction that ranks them according to the costs incurred. The proposed adaptive framework has been integrated into\nthe Vienna Cloud Environment (VCE) and the evaluation by means of a data-intensive application is presented herein....
In April 2012, Rogers and Cliff (R&C) demonstrated a theoretical financial brokerage model for cloud computing that is\nprofitable for the broker, offers reduced costs for cloud users, and generates more predictable demand flow for cloud\nproviders. Relatively cheap, long-term reserved instances (RIs) are bulk-purchased by the broker, and then\nre-packaged and re-sold as monthly options contracts at a price lower than a user can purchase ââ?¬Å?on-demandââ?¬Â from\nthe provider. Thus, the broker risks exposure on purchase for margin on sales. R&Cââ?¬â?¢s result has generated significant\ninterest in the cloud computing community and is currently the fifth most accessed research paper of all time in the\nJournal of Cloud Computing: Advances, Systems and Applications.\nHere, we perform an independent replication of R&Cââ?¬â?¢s brokerage model using CReST, a discrete event simulation\nplatform for cloud computing developed at the University of Bristol. We identify two implementation problems in\nR&Cââ?¬â?¢s original work: firstly, the broker buys fewer RIs than the model suggests; and secondly, the broker is\nundercharged for RIs used. We correct R&Cââ?¬â?¢s results accordingly: while brokerââ?¬â?¢s profits are not as high as R&C suggest,\nthe model still supports the theoretical possibility of a profitable brokerage.\nHowever, aggressive competition between cloud providers has reduced the cost of cloud services to users and led to\nthe introduction of new secondary markets where users can buy and sell RIs between themselves. This has squeezed\nthe opportunity for an intermediary brokerage. By recalibrating R&Cââ?¬â?¢s model to fit current market conditions, we\nconclude that the commercial viability of R&Cââ?¬â?¢s brokerage model has been eradicated. The window of opportunity\nhas now closed....
In this paper, we describe the development of template management technology to build virtual resources\nenvironments on OpenStack. In recent days, Cloud computing has been progressed and also open source Cloud\nsoftware has become widespread. Authors are developing cloud services using OpenStack. There are technologies\nwhich deploy a set of virtual resources based on system environmental templates to enable easy building,\nexpansion or migration of cloud resources. OpenStack Heat and Amazon CloudFormation are template deployment\ntechnologies and build stacks which are sets of virtual resources based on templates. However, these existing\ntechnologies have 4 problems. Heat and CloudFormation transaction managements of stack create or update are\ninsufficient. Heat and CloudFormation do not have sharing mechanism of templates. Heat cannot extract templates\nfrom existing virtual environments. Heat does not reflect actual environment changes to stack information.\nTherefore, we propose a new template management technology with 4 improvements. It has a mechanism of\ntransaction management like roll back or roll forward in case of abnormal failure during stack operations. It shares\ntemplates among end users and System Integrators. It extracts templates from existing environments. It reflects\nactual environment changes to stack information. We implemented the proposed template management server\nand showed that end users can easily replicate or build virtual resources environments. Furthermore, we measured\nthe performance of template extraction, stack create and update and showed our method could process templates\nin a sufficient short time....
Cloud service abstractions are currently used to hide the underlying complexity given by existing technologies and\nservices, in hope of facilitating the enacting of Cloud Federations and Marketplaces. In particular, resource\nmanagement systems dealing with multiple Cloud providers need to expose an uniform interface for various services\nand to build wrappers for the Cloud service APIs. In this paper we discuss the solution adopted by a recent developed\nopen-source and vendor agnostic platform-as-a-service for Multi-Cloud application deployment. The middleware\nincludes a multi-agent system for automatic Cloud resource management. With a modular design, the solution\nprovides a flexible approach to encompass new Cloud service offers as well as new resource types. This paper focuses\non the modules which enable resource abstraction and automatized management....
Loading....