Center for Distributed Object Computing

Director: Chris Gill

Home Page: http://128.252.165.3/~schmidt/research.html Location: 506 Bryan Hall Email: cdgill@wustl.edu

Research at the Center for Distributed Object Computing at Washington University focuses on design patterns, implementation, and experimental analysis of object-oriented techniques that facilitate the development of high-performance, real-time distributed object computing frameworks on parallel processing platforms running over high-speed networks. The primary goal of this center is to support advanced R&D on distributed object computing middleware using an open source software development model. This model allows academics, developers, and end-users to participate in leading-edge R&D projects driven by the free market of ideas, requirements, and resources.

News Stories & Tip Sheets:

Showing 1 Stories.

Free networking opportunities State-of-the-art laboratory offered to researchers, students Sept. 7, 2005 -- A router in the new Open Network Laboratory, funded by NSF. A novel networking service has been made available to the research community by computer scientists at Washington University in St. Louis, enabling researchers and students remote, free use of the latest networking technology. Ultimately, the new Open Network Laboratory (ONL )can lead to innovations that can expand the capability of the Internet and other networking environments, said its director, Jonathan S. Turner, Ph.D., Henry Edwin Sever Professor of Engineering, and professor of computer science and engineering at WUSTL.

Showing 1 Stories.

Faculty Experts:

Motivation for the Center on Distributed Object Computing

The emergence of the Internet is the most important technological event in the last decade of the 20th century. It is becoming pervasive and will influence research, business, and personal life well into the 21st century. Its growth is phenomenal and it is stretching its original design goals. It must be the next "trade route,'' the next "marketplace,'' the next "information pipeline.'' The Next Generation Internet (NGI) must have a whole new set of characteristics and powerful features. Most importantly, it should be based on standard, completely open middleware.

Middleware is software that users do not see, but it is key to openness and continued innovation through competition. Its role is to broker the communication between consumers and suppliers. Middleware masks differences between OS platforms and networks. For instance, clients can request a service without knowing where or how that service is implemented. This flexibility will become even more important because the NGI architecture is based on the fundamental premise that unlike today's computer-centric network, tomorrow's Internet will be a heterogeneous mix of cooperative, intelligent devices located in the home, at the factory, in public places, and at the office. These network devices will not just be PCs, they will increasingly be intelligent home appliances and sensors, personal communication devices, entertainment centers, and other novel forms of electronics.

For the Internet to become pervasive, access devices must become very cheap, their use must be intuitive, and distributed computing middleware must do most of the "thinking'' for the users. Thus, today's network architectures, where applications rely largely on local operating system services, must evolve. This evolution will turn the current computing model inside out -- clients will become very thin, the intelligence will reside in the network, and most applications will be network-centric. Resources can then be shared, to reduce their cost and facilitate maintenance and updates.

Intelligent middleware is the key enabling technology to realize the NGI vision. An open standard by the OMG called CORBA is rapidly gaining mindshare and marketshare in both the research and commercial domains. Unfortunately, first-generation CORBA ORBs did not provide adequate quality of service (QoS) for performance-sensitive NGI applications, such as teleconferencing and Internet telephony, because requests were treated with "best effort'' response. All NGI applications are not created equal, however. Some must run faster, more reliably and more consistently, e.g., in a predictable time context, than others. In a heterogeneous environment, systems and networks can provide better QoS by prioritizing services. As network providers broaden their bandwidth and service offerings, flexibility becomes increasingly more important, since pricing must be sensitive to both performance and "class of service'' if it is to reach the widest possible market.

The DOC Group's Contribution to R&D on Distributed Object Computing

Since 1995, the DARPA Quorum program, the National Science Foundation (NSF), and many major industrial sponsors have funded the Distributed Object Computing (DOC) group to conduct leading-edge software research on the characteristics and features required for the NGI. The primary contributions of this research have been two key middleware software packages that can support QoS for the NGI -- ACE and TAO:

• The ADAPTIVE Communication Environment (ACE) is a freely available open source object-oriented (OO) framework that implements many core design patterns for concurrent communication software. ACE provides a rich set of reusable C wrappers and framework components that perform common communication software tasks across a broad range of OS platforms. The communication software tasks provided by ACE include event demultiplexing and event handler dispatching, signal handling, service initialization, interprocess communication, shared memory management, message routing, dynamic (re)configuration of distributed services, concurrent execution and synchronization.

• "The ACE ORB'' (TAO) is a CORBA-compliant Object Request Broker (ORB) that addresses the QoS requirements of NGI applications andprovides many standard CORBA services. Like ACE, TAO is freely available using the open source model. Moreover, it leverages ACE components to communicate efficiently, predictably, and portably with operating systems and network protocols on the application's behalf to establish the necessary end-to-end QoS. Empirical performance analysis indicates that TAO's second-generation architectural design has yielded a high-performance, real-time ORB with highly predictable response over a wide range of demanding environments. For instance, TAO contains the first real-time ORB endsystem that supports end-to-end QoS guarantees over high-speed ATM networks and embedded system interconnects.

Our research on ACE and TAO is motivated by the recognition that advances in communication middleware software can be achieved only by simultaneously investigating techniques, patterns, and tools that (1) simplify software development, (2) optimize system performance, and (3) rigorously measure system behavior to pinpoint and alleviate performance bottlenecks and sources of priority inversion and non-determinism.

In addition to working closely with our research sponsors on ACE and TAO, many other companies have deployed our middleware in commercial projects and research labs around the world, where it is used for a broad spectrum of systems ranging from telecommunications, medical imaging, avionics, simulation, and financial services. As a testament to our success in technology transfer, a independent company, Riverace, has formed to provide commercial support for ACE and a St. Louis company, OCI, has recently begun to provide commercial support for TAO.