http://hdl.handle.net/1880/45343 2013-05-26T03:26:19Z 2013-05-26T03:26:19Z Xiao, Zhonge Unger, Brian W http://hdl.handle.net/1880/46083 2012-03-30T20:04:40Z 1998-07-01T00:00:00Z

Title: Report on WarpKit performance study and improvement Authors: Xiao, Zhonge; Unger, Brian W Abstract: This is a report on the earlier development of WarpKit, a parallel simulation kernel based on shared-memory multi-processor architecture, as part of the Telesim project. The development is aimed at exploiting shared memory multi- processor paradigm and developing a Parallel Discrete Event Simulation package which is based on shared memory multi-processors and capable of delivering high performance. Three major problems that have great impact on the performance of Time Warp systems are: excessive cost incurred by rollback computation resulting from sole reliance on rollback as a basic synchronization mechanism in a distributed/parallel processing system, large amount of memory space required to run applications, and high system overheads in inter-process communication and global control (e.g. GVT computation and memory management). Shared memory multi-processor architecture provides the potential of delivering much higher performance for Time Warp systems than can be achieved in distributed environment. New approaches could be conceived to address these problems and to realize the potential. This report covers the results of our effort to improve WarpKit Kernel performance. Incremental State Saving has been implemented on top of existing Kernel which reduces both the time and space spent on state saving, a necessity of Time Warp. Purely asynchronous schemes have been developed and implemented for the global control mechanism. As a result, the system overhead on global control has been reduced significantly. The new global control mechanism also makes the system overheads insensitivity to the number of processors as opposed to the distributed situation where system overhead experiences a sharp increase with the number of processors. A global scheduling and load balancing mechanism is expected to restrict the number of rollbacks to a low percentage over net events to be processed by the Kernel. With these new mechanisms in place, one may expect close to linear speedup curve for parallel discrete event simulation on shared memory multi- processors.

1998-07-01T00:00:00Z Unger, Brian W http://hdl.handle.net/1880/46079 2012-03-30T16:20:57Z 1988-03-01T00:00:00Z

Title: Distributed software via prototyping and simulation - JADE Authors: Unger, Brian W Abstract: Jade is an environment that supports the development of distributed software. Components may be written in any of a number of different languages, with a common inter-process communication protocol providing a uniform interface among the components. A window system allows the user to interact with many different processes at once. A hierarchical graphics system is provided for use with documentation and programming, and for support of monitoring. Monitoring in Jade is also supported by an extensible mechanism which allows for multiple views of the same process. The nondeterminism of distributed systems may be controlled in order to provide repeatability of executions and to aid in testing and debugging. Finally, the formal specification of inter-process events in Jade is supported by a communications protocol verifier, allowing run-time consistency checking. Together, these tools provide a powerful environment for software prototyping and simulation. This paper is a summary of work that has been described in [Unger 85], [Lomow 85] and [Joyce 87].

1988-03-01T00:00:00Z Li, Xining Unger, Brian W http://hdl.handle.net/1880/46078 2012-03-30T17:17:32Z 1988-01-01T00:00:00Z

Title: Predicting X-tree network performance using the JADE environment Authors: Li, Xining; Unger, Brian W Abstract: Jade provides an integrated set of tools which are designed to support the development of distributed software and systems. The jade environment provides tools for the design, implementation, debugging, testing, maintenance, simulation, and performance analysis of distributed, concurrent programs. A network topology called X-tree has been implemented and simulated using this Jade environment. This paper presents an overview of the Jade environment, the X-tree network topology and a robust routing algorithm for this topology. The performance of the X-tree topology is also discussed.

1988-01-01T00:00:00Z Li, Xining Unger, Brian W Cleary, John Lomow, Greg West, Darrin http://hdl.handle.net/1880/46077 2012-03-30T20:54:10Z 1988-01-01T00:00:00Z

Title: Communicating Sequential Prolog Authors: Li, Xining; Unger, Brian W; Cleary, John; Lomow, Greg; West, Darrin Abstract: Communicating Sequential Prolog (CSP') is a single-solution distributed logic programming language for discrete event simulation. Its primary goal is to speed up the execution of logic programs through the use of parallelism, while as far as possible preserving the semantics of standard Prolog. A CSP' program consists of a set of parallel processes, synchronized by simulation time and by message passing. The underlying interprocess communication mechanism is Time Warp. The pertinent features of parallel logic programming and Time Warp are described. The syntax of CSP' is introduced and the semantics of new predicates and their control structures are discussed. Examples are given to show the expressive power and simplicity of CSP'.

1988-01-01T00:00:00Z