Opening Keynote

Michael Franklin
University of California, Berkeley

Title: Dataspaces: Progress and Prospects

Abstract

The concept of Dataspaces was introduced in late 2005 by Franklin, Halevy, and Meier in an attempt to articulate a new data management paradigm, one that builds on much of the work that is already going on in the database community. The key is to move away from the schema-first nature of traditional data integration techniques. Instead, Dataspaces offer an initial set of useful services over a set of data sources, allowing users, administrators, or programs to improve the semantic relationships between the data in a pay-as-you-go manner. Dataspace techniques enable rapid deployment of integrated systems while enabling reconciliation and structuring efforts to be applied only in those cases where the benefit of doing so is apparent. In this talk, I will outline the motivation behind Dataspaces, describe the evolution of thinking on the topic since the 2005 proposal, and present recent and ongoing work in the area.

Biographical information

Michael Franklin is a Professor of Computer Science at UC Berkeley where his research focuses on the architecture and performance of distributed data management and information systems. His recent projects cover the areas of data stream processing, large-scale sensor networks, high-speed message brokering, cloud and grid computing, and data and application integration. He is also CTO of Truviso, a Data Analytics company he co-founded in 2006.

After a five-year stint as a database systems developer he attended the University of Wisconsin, Madison, where he received his Ph.D. in Computer Science in 1993. He was on the faculty at the University of Maryland before joining Berkeley in 1999. He is currently on the editorial boards of the ACM Journal of Data Quality and the VLDB Journal and is a trustee and member of the Executive Committee of the VLDB endowment. He served as Conference Co-Chair for the CIDR 2009 conference and is Co-Chair of the Industrial Track for SIGMOD 2009. He is a Fellow of the Association for Computing Machinery and is a recipient of the National Science Foundation CAREER Award and the ACM SIGMOD "Test of Time" award.

Closing Keynote

Domenico Laforenza
Italian National Research Council (CNR)

Title: XtreemOS: Towards a Grid-enabled Linux-based Operating System

Abstract

Since the term "Grid Computing" was introduced at the end of the 90s by Foster and Kesselman, several researchers have proposed to revise the initial definition of Grid. More recently researchers at the European Network of Excellence "CoreGrid" reached an agreement on the following definition: a Grid is "a fully distributed, dynamically reconfigurable, scalable and autonomous infrastructure to provide location independent, pervasive, reliable, secure and efficient access to a coordinated set of services encapsulating and virtualizing resources (computing power, storage, instruments, data, etc.) in order to generate knowledge".

This is a more modern service-oriented vision of the Grid that stems from the conviction that in the mid to long term the great majority of complex software applications will be dynamically built by composing services, which will be available in an open market of services and resources. In this sense, the Grid will be conceived as a "world-wide cyber-utility" populated by cooperating services interacting as in a complex and gigantic software ecosystem.

While Grid Computing has gained much popularity over the past few years in the scientific contexts, it is still cumbersome to use effectively in business and industrial environments. In order to help remedy this situation, several researchers proposed to build a true Grid Operating System (GOS). A GOS is a distributed operating system targeted for a large-scale dynamic distributed architecture, with a variable amount of heterogeneous resources (resources may join, leave, churn). A GOS should be aware of Virtual Organizations (VO), spanning multiple administrative domains with no central management of users and resources (multiple administrators, resource owners, VO managers). This GOS should be composed of a consistent set of integrated system services, providing a stable Posix-like interface for application programmers. Moreover, abstractions or jobs (set of processes), files, events, etc. should be provided by a GOS.

This talk will present XtreemOS, a first European step towards the creation of a true open source operating system for Grid platforms. The XtreemOS project aims to address this challenge by designing, implementing, experimenting with and promoting an operating system that will support the management of very large and dynamic ensembles of resources, capabilities and information composing virtual organizations. Recognizing that Linux is today the prevailing operating system, XtreemOS started an effort to extend Linux towards Grid, including a native support for the VOs management, and providing appropriate interfaces to the GOS services. As it will be explained in this talk, the result is neither a "true" Grid operating system nor a Grid middleware environment, but rather a Linux operating systems with tightly integrated mechanisms for the quick and user-friendly creation of distributed collaborations which share their resources in a secure and user-friendly way.

Biographical information

Dr. Domenico Laforenza is Director of the Institute for Informatics and Telematics, belonging to the Italian National Research Council (CNR), since the 1st of July 2008. He is also head of the Registry of the country code Top Level Domain .it (ccTLD .it).

Since the June 2008 he was responsible for the High Performance Computing Laboratory at the Information Science and Technologies Institute (ISTI) of the Italian National Research Council (CNR). Dr. Laforenza received the doctoral degree in Computer Science from the Department of Computer Science of University of Pisa in 1977. His main fields of research interest are high performance computing and the programming of distributed and parallel systems. In the past (2002-2003) he was member of the Grid Forum Steering Committee (GFSC) and co-Director of the Grid Performance and Information Services Area. Dr. Laforenza contributed to the creation of CoreGRID (EU-IST-FP6-004265), the European Research Network on Foundations, Software Infrastructures and Applications for large scale distributed, GRID and Peer-to-Peer Technologies, funded by the European Commission's 6th Framework Program. From September 2004 to August 2008 he was chairman of the CoreGrid Members General Assembly, and member of the Integration Monitoring Committee and the Executive Committee. In June 2006 he was appointed as chairman of the Governing Board of the IST FP6 XtreemOS project (EU IST-FP6-033576).

Currently he is actively involved in the European Network of Excellence "S-CUBE" (EU IST-FP7-215483).

He is also member of the Scientific Advisory Board of the Information and Communication Technology (ICT) Department of the Italian National Research Council (CNR), and since January 2008 he is member of the Scientific Advisory Board of the Institute de Grilles CNRS.

More information at: http://miles.isti.cnr.it/~lafo/domenico/domenico.html