Ian Kelley

Optimizing the data distribution layer of BOINC with BitTorrent

In this paper we show how we applied BitTorrent data distribution techniques to the BOINC middleware. Our goal was to decentralize BOINCs data model to take advantage of client network capabilities. To achieve this, we developed a prototype that adds BitTorrent functionality for task distribution and conducted small- scale tests of the environment. Additionally, we measured the impact of the BitTorrent components in both the BOINC client and server, and compared it with the original implementation. Our preliminary results indicate that the BitTorrent client had a negligible influence on the BOINC clients computation time, even in the case where it was seeding extensively. The BOINC server, on the contrary, showed an unexpectedly low bandwidth output when seeding the file, as well as spikes on CPU usage. Current results show the BitTorrent scenario allows clients to share the burden of data distribution on BOINC with almost no negative influence on compute time. This paper will discuss the tests that were performed, how they were evaluated, as well as some improvements that could be made to future tests to enhance server-side efficiency.

OPTIMIZING DATA DISTRIBUTION IN DESKTOP GRID PLATFORMS

Current infrastructures for Volunteer Computing follow a centralized architecture for data distribution, creating a potential bottleneck when tasks require large input files or the central server has limited bandwidth. In this paper we propose two new data models for Berkeley Open Infrastructure for Network Computing (BOINC): an approach based on the popular BitTorrent protocol; and a Content Delivery Network approach. While the latter remains on a theoretical level, we developed a prototype that adds BitTorrent functionality for task distribution and conducted medium-scale tests of the environment. Our preliminary results indicate that the BitTorrent client had a negligible influence on the BOINC clients computation time. The BOINC server showed an unexpectedly low bandwidth output when seeding the file, as well as spikes on CPU usage. This paper discusses the tests that were performed, how they were evaluated, as well as some improvements that could be made in future research on both approaches.

EDGeS: The common boundary between service and desktop grids

Service grids and desktop grids are both promoted by their supportive communities as great solutions for solving the available compute power problem and helping to balance loads across network systems. Little work, however, has been undertaken to blend these two technologies together. In this paper we introduce a new EU project, that is building technological bridges to facilitate service and desktop grid interoperability. We provide a taxonomy and background into service grids, such as EGEE and desktop grids or volunteer computing platforms, such as BOINC and XtremWeb. We then describe our approach for identifying translation technologies between service and desktop grids. The individual themes discuss the actual bridging technologies employed and the distributed data issues surrounding deployment.

Bridging the Data Management Gap Between Service and Desktop Grids

Volunteer computing platforms have become a popular means of providing vast amounts of processing power to scientific applications through the use of personal home computers. To date, with little exception, these systems have focused solely on exploiting idle CPU cycles and have yet to take full advantage of other available resources such as powerful video card processors, hard disk storage capacities, and high-speed network connections. As part of the EDGeS project, we are working to expand this narrow scope to also utilize available network and storage capabilities. In this paper we outline the justifications for this approach and introduce how decentralized P2P networks are being built in the project to distribute scientific data currently on the Grid.

Peer-To-Peer Techniques for Data Distribution in Desktop Grid Computing Platforms

In this paper, we discuss how Peer-to-Peer data distribution techniques can be adapted to Desktop Grid computing environments, particularly to the BOINC platform. To date, Desktop Grid systems have focused primarily on utilizing spare CPU cycles, yet have neglected to take advantage of client network capabilities. Leveraging client bandwidth will not only benefit current projects by lowering their overheads but also will facilitate Desktop Grid adoption by data-heavy applications. We propose two approaches to Peer-to-Peer data sharing that could be adapted for volunteer computing platforms: the highly successful BitTorrent protocol and a secure and customizable Peer-to-Peer data center approach.

A peer-to-peer architecture for data-intensive cycle sharing

It has been nearly a decade since volunteer computing was first popularized by the highly successful SETI@Home project. SETI@Homes core software evolved and became the Berkeley Infrastructure for Open Network Computing (BOINC). This development paved the way for many other scientific projects to leverage the power of volunteer computing and Desktop Grids. More recently, there has been a movement to transition many of the applications that run at large-scale computing facilities to Desktop Grid platforms. One of the challenges that arises when moving jobs from Service Grids to Desktop Grids is managing data distribution to the numerous and highly distributed worker nodes. In this paper we present a data sharing architecture that enhances file sharing within Desktop Grids while also easing Service to Desktop Grid migration. The architecture presented here takes into consideration many of the specific needs of scientific volunteer computing applications, such as control of security aspects, legacy application integration, and mechanisms for participants to opt-in and opt-out of the system.

EDGeS: A Bridge between Desktop Grids and Service Grids

Desktop grids and service grids widely used by their different users communities as efficient solutions for making full use of computing power and achieving loads balances across Intranet or Internet. Nevertheless,little work has been done to combine these two grids technologies together to establish a seamless and vast grid resources pool. In this paper we will present a new European FP7 infrastructure project:EDGeS (enabling desktop grids for e-science), which aim to build technological bridges to facilitate interoperability between desktop grid and service grid. We give also a taxonomy of existing grid systems: desktop grids such as BONIC and XtremWeb, service grids such as EGEE. Then we describe furtherly our solution for identifying translation technologies for porting applications between desktop grids and service grids, and vice versa. There are three themes in our solution, which discuss actual popular bridging technologies, user access issues, and distributed data issues about deployment and application development.

Dynamic service-based integration of mobile clusters in grids

The emergence of pervasive and mobile computing has drawn research attention to integrated mobile Grid systems. These new hybrid Grids consist of a typical SOA backbone extended to include mobile and small scale devices such as personal digital assistants, smart-phones, multimedia devices, and intelligent sensors. In a fully integrated model, mobile devices are able to act both as consumers and providers to open up a completely new range of very interesting possibilities in exploiting their mobile nature, unique functionality, and context awareness. However, in resource-limited environments, traditional SOA frameworks cannot easily be deployed since they assume a plethora of available device resources, and have a number of complex dependencies, thus rendering them unsuitable for resource-constrained devices. Therefore, a smaller and simpler server-side container with reduced requirements and dependencies is needed. The contribution of this paper is two-fold: first, we have designed a J2ME-compliant socketbased server-side container, and second, we have demonstrated how an aggregator framework enables such mobile services to be accessed using standard-based Web services in a high-level manner.

Porting applications to a Distributed Computing Infrastucture incorporating Desktop Grids

The European EDGeS (Enabling Desktop Grids for e-Science) and its follow-up project EDGI (European Desktop Grid Initiative) have supported several large European user communities from diverse application areas in utilizing Desktop Grid resources for their computation. DEGISCO (Desktop Grids for International Scientific Collaboration), another FP7 project supports several non-European user communities in utilizing this infrastructure. Both projects use a generic methodology called EADM (EDGeS Application Development methodology) when porting applications to a combined Service Grid/Desktop Grid platform.