Middleware Support for Global Access to Integrated Computational Collaboratories

 

Overview

 

A Collaboratory is defined as a place where scientists and researchers work together to solve complex interdisciplinary problems, despite geographic and organizational boundaries. Computational collaboratories provide uniform (collaborative) access to computational resources, services and/or applications. These systems expand the resources available to researchers, enable multidisciplinary collaborations and problem solving, increase the efficiency of research, and accelerate the dissemination of knowledge.

 

The growth of the Internet and the advent of the computational “Grid” have made it possible to develop and deploy advanced computational collaboratories. Recent efforts include the Upper Atmospheric Research Collaboratory (UARC), Diesel Combustion Collaboratory (DCC), Access Grid, RCSB, EMSL, Cactus, Astrophysics Simulation Collaboratory and DISCOVER.  Each of these systems provides a high-level problem-solving environment  (PSE) that builds on the underlying Grid technologies to provide seamless access to domain specific resources, services and applications. Together these systems have the potential for enabling truly global scientific investigation through the creation of meta-laboratories spanning many research groups, universities and countries, and transforming computational applications and simulations into global modalities for research and instruction.

 

While combining these systems can lead to truly collaborative, multi-disciplinary and multi-institutional problem solving, integrating these  “focused” collaboratories presents significant challenges. This is because each of these systems has a unique architecture and implementation, and builds on different enabling technologies. Key among these challenges is the design and development of robust middleware support that addresses scalability, service discovery, security and access control, and interaction and collaboration management for consistent access.  Such a middleware should define a minimal set of interfaces and protocols to enable collaboratories to share resources, services and applications on the Grid while being able to maintain their architectures and implementations of choice.

 

This research initiative investigates the architecture of such a middleware that enables integration of and global access to computational collaboratories.  We present the design and implementation of a middleware substrate that enables a peer-to-peer integration of and global collaborative web-based access to multiple, distributed instances of the DISCOVER computational collaboratory. DISCOVER provides collaborative access to high-performance parallel and distributed applications for interaction and steering using web-based portals. The key design challenge is enabling scalable, secure, consistent and controlled access to remote, highly dynamic distributed applications for real-time monitoring, interaction and steering by geographically distributed scientists and engineers in a collaborative environment. The middleware substrate enables DISCOVER interaction and steering servers to dynamically discover and connect to one another to form a peer-to-peer network. This allows clients connected to their local servers to have global access to all applications and services across all the servers in the network based on their credentials, capabilities and privileges. The design and implementation of the DISCOVER middleware substrate builds on existing web servers and leverages commodity technologies and protocols such as CORBA and HTTP. Its goal is to enable rapid deployment, ubiquitous and pervasive access, and easy integration with 3rd party services, while evaluating the viability of these technologies for advanced Grid applications.

 

The overall aim of Grid computing is to enable collaborative and coordinated problem solving in dynamic, multi-institutional virtual organizations and it focuses on large-scale resource sharing, innovative applications, and high performance computing. The middleware substrate addresses one aspect of this general problem by providing global collaborative access to grid applications and services.

 

We also try to present a retrospective evaluation of our design and discuss its advantages and disadvantages. An evaluation of commodity distributed technologies and protocols and their ability to support Grid applications is also presented and some open issues and challenges are discussed.

 

 

Publications

 

[1].         V. Mann and M. Parashar, Middleware Support for Global Access to Integrated Computational Collaboratories, accepted for publication and presentation at the Tenth IEEE Symposium on High Performance Distributed Computing (HPDC-10), San Francisco, CA, August 2001. (PDF)

[2].         S. Kaur, V. Mann, V. Matossian, R. Muralidhar, M. Parashar, "Engineering a Distributed Computational Collaboratory", 34th Hawaii Conference on System Sciences, January 2001. (PDF)