Cloud Computing has become another buzzword after Web 2.0. However, there are dozens of different definitions for Cloud Computing and there seems to be no consensus on what a Cloud is. On the other hand, Cloud Computing is not a completely new concept; it has intricate connection to the relatively new but thirteen-year established Grid Computing paradigm, and other relevant technologies such as utility computing, cluster computing, and distributed systems in general. This paper strives to compare and contrast Cloud Computing with Grid Computing from various angles and give insights into the essential characteristics of both.

Using virtual machines as a resource provisioning mechanism offers multiple benefits, most recently exploited by “infrastructure-as-a-service ” clouds, but also poses several scheduling challenges. More specifically, although we can use the suspend/resume/migrate capability of virtual machines to support advance reservation of resources efficiently, by using suspension/resumption as a preemption mechanism, this requires adequately modeling the time and resources consumed by these operations to ensure that preemptions are completed before the start of a reservation. In this work we present a model for predicting various runtime overheads involved in using virtual machines, allowing us to efficiently support advance reservations. We extend our lease management software, Haizea, to use this new model in its scheduling decisions, and we use Haizea with the OpenNebula virtual infrastructure manager so the scheduling decisions will be enacted in a Xen cluster. We present both physical and simulated experimental results showing the degree of accuracy of our model and the long-term effects of variables in our model on several workloads. 1

Abstract. Compute-intensive biological applications are heavily reliant on the availability of computing resources. Grid based HPC clusters and emerging Cloud computing clusters provide a large scale computing environment for scientific users. However, large scale biological application often involves various types of computational tasks which can benefit from different types of computing clusters. Therefore, a high level job scheduling environment which integrates the Grid style HPC clusters and the Cloud computing clusters and manages jobs accordingly based on the characteristics of the jobs is required. In this paper, we propose a Web service framework for high-level job scheduling – Swarm. Swarm is developed for scientific applications that must submit massive number of high-throughput jobs or workflows to highly distributed computing clusters. Swarm allows the users to submit jobs to both Grid HPC and Cloud computing clusters. The Swarm service itself is designed to be extensible, lightweight, and easily installable on a desktop or a small server. As a Web service, derivative services based on Swarm can be straightforwardly integrated with Web portals and science gateways. This paper provides the motivation for this research, the architecture of the Swarm framework, and a performance evaluation of the system prototype.

Abstract—Efficient data management is a key component in achieving good performance for scientific workflows in distributed environments. Workflow applications typically communicate data between tasks using files. When tasks are distributed, these files are either transferred from one computational node to another, or accessed through a shared storage system. In grids and clusters, workflow data is often stored on network and parallel file systems. In this paper we investigate some of the ways in which data can be managed for workflows in the cloud. We ran experiments using three typical workflow applications on Amazon’s EC2. We discuss the various storage and file systems we used, describe the issues and problems we encountered deploying them on EC2, and analyze the resulting performance and cost of the workflows. Index Terms—Cloud computing, scientific workflows, cost evaluation, performance evaluation.

We present our ideas for modeling groups of interdependent virtual machines in the cloud. We call these models virtual environments. This abstraction is built on top of virtual appliances and the services they provide. We discuss previous attempts in this domain and present our motivations for working on an uncomplicated model for non-expert users of cloud computing such as Web developers and CS students. Visual and internal representations of the model are presented. Early work on a prototype implementation is described. We argue that easier to use models such as ours are needed for today’s and tomorrow’s distributed applications.

The fast development of cloud computing systems stimulate the needs for a standalone block storage system, which can provide persistent block storage services to the virtual machines maintained by the clouds. This paper presents the Virtual Block Store (VBS) System, a standalone block storage system built on the basis of LVM, iSCSI, and Xen hypervisor, which can provide basic block storage services such as volume creation and attachment. The concept and functional interface of VBS are based on the Amazon Elastic Block Store (EBS) service; moreover, VBS can be used independently with an existing volume server and Xen nodes, and can be easily extended to support other virtual machine managers, or integrated with various cloud computing systems. Preliminary experiments show that a VBS volume can provide I/O performance that is similar to an ATA over Ethernet virtual device, and comparable to a local logical volume. 1.

Revision 33 NOTE: I passed my candidacy exam on 11/25/08. At that time, my committee suggested some changes to my proposal (which are not included in this document). Nonetheless, this

Currently, e-Science applications run on dedicated computing platforms. The scientific/e-Science computing community developed the Grid to share resources and maximize compute capacity available for e-Science applications. Applications exist that incidentally have urgent high demand for compute capacity, for example in case of emergencies. Also, some researches need to do experiments with Grid computing itself. For such cases, the Grid is not ideally equipped. Recently Cloud computing emerged, providing on demand compute resources on a pay-per-use basis. Clouds promise infinite virtual resources. This report addresses the utilization of (additional) resources offered by Clouds for Grid applications. To this end, a working implementation of a Grid enabled Cloud compute cluster is presented and assessed.

Abstract – Cloud Computing has become another buzzword after Web 2.0. However, there are dozens of different definitions for Cloud Computing and there seems to be no consensus on what a Cloud is. On the other hand, Cloud Computing is not a completely new concept; it has intricate connection to the relatively new but thirteen-year established Grid Computing paradigm, and other relevant technologies such as utility computing, cluster computing, and distributed systems in general. This paper strives to compare and contrast Cloud Computing with Grid Computing from various angles and give insights into the essential characteristics of both. 1 100-Mile Overview Cloud Computing is hinting at a future in which we won’t compute on local computers, but on centralized facilities operated by third-party compute and storage utilities. We sure won’t miss the shrink-wrapped software to unwrap and install.

Abstract — The commercial success of Cloud computing and recent developments in Grid computing have brought platform virtualization technology into the field of high performance computing. Virtualization offers both more flexibility and security through custom user images and user isolation. In this paper, we deal with the problem of distributing virtual machine (VM) images to a set of distributed compute nodes in a Cross-Cloud computing environment, i.e., the connection of two or more Cloud computing sites. Ambrust et al. [3] identified data transfer bottlenecks as one of the obstacles Cloud computing has to solve to be a commercial success. Several methods for distributing VM images are presented, and optimizations based on copy on write layers are discussed. The performance of the presented solutions and the security overhead is evaluated. I.