Using Schooner to support distribution and heterogeneity in the Numerical Propulsion System Simulation project

Abstract

SUMMARYThe Numerical Propulsion System Simulation (NPSS) project has been initiated by NASA to explore the use of computer simulation in the development of new aircraft propulsion technology. With this approach, each engine component is modeled by a separate computational code, with a simulation executive connecting the codes and modeling component interactions. Since each code potentially executes on a different machine in a network, a simulation run is a heterogeneous distributed program in which diverse software and hardware elements are incorporated into a single computation. In the paper a prototype simulation executive that supports this type of programming is described. The two components of this executive are the AVS scientific visualization system and the Schooneir heterogeneous remote procedure call (RPC) facility. In addition, the match between Schooner's capabilities and the needs of NPSS is evaluated based on our experience with a coUectioii of test codes. The basic conclusion is that, while Schooner fared well in general, it exhibited certain deficiencies that dictated changes in its design and implementation. This discussiam not only documents the evolution of Schooner, but also serves to highlight the practical problems that can be encountered when dealing with heterogeneity and distribution in such applications.

Interconnecting interactive and remote parallel components using schooner

“…To support this symmetric strategy, Schooner provides the abstraction of lines, where each line includes a thread of control and naming scope analogous to a normal Schooner metacomputation. 5 Lines allow the type of independent communication needed here, but still limit interaction patterns relative to arbitrary message passing to preserve ease of use and implementation simplicity. A symmetric strategy based on lines is currently being used in the ecosystem application described in section 2.1 to implement independent communication between processes in a parallel simulation and separate components containing a parallel Geographical Information System (GIS) and custom visualization tool, respectively.…”

“…This approach involves incorporating one or more parallel computations and a scientific visualization system such as AVS 2 into a single heterogeneous distributed program or metacomputation 3 using the Schooner software interconnection system. [4][5][6] With this scheme, Schooner is used to transfer data and control transparently with Remote Procedure Call (RPC) semantics between the visualization system executing on a workstation and parallel computations running on a variety of remote platforms. The experiments reported here have involved AVS and a parallel computation running on either an Intel Paragon, a Sequent Symmetry, or a collection of Sun Sparcstations using PVM 7,8 (Figure 1).…”

A Scalable Distributed Multimedia Knowledge Retrieval System on a Cluster of Heterogeneous High Performance Architectures

“…In order to get a simulation of a complex object in realistic time, the different components of the object have to be simulated on different clusters of computers to exploit component level parallelism [14], and the results of the multiple simulations have to be reconciled. Since the architectures of high performance computers are very different and are continuously changing, it is necessary to design a simulation system which is architecture independent, supports high performance clusters of heterogeneous architectures, is scalable, and is portable.…”

A distributed multimedia knowledge based environment for modeling over the Internet

“…We are extending our system to provide interfaces to already developed libraries of numerical codes in NPSS project to simulate an aircraft combustion engine [5].…”

“…This system aims to develop an user friendly environment for the analysis and design of aircraft engines by integrating existing numerical codes for engine components [2,5,9] to create an end-to-end engine simulation in realistic time. In this paper, we will use examples from NASA engine simulation project NPSS [5] to illustrate our system.…”