Xiaoge Wang

Design of an OGSA Based MetaService Architecture

This paper proposes a MetaService-based architecture for dealing with the problem of services management in grid systems. We introduce the basic idea of MetaService, and then discuss MetaService classification, new MetaService construction, and the MetaService working mechanism. A graph and a tree can be used for the optimization function of MetaService. An example from real life is used to explain the role of MetaService. We give an overview of a grid prototype system which has incorporated some of the ideas presented. A brief comparison is given between MetaService and other related work.

MyVIA: a design and implementation of the high performance Virtual Interface Architecture

Virtual Interface Architecture (VIA) established a communication model with low latency and high bandwidth, and defined the standard of user-level high-performance communication specification in cluster systems. This paper analyzes the current development, principle and implementations of VIA, and presents user-level high-performance communication software, MyVIA, based on Myrinet, which is comfortable with VIA specification. The paper first describes the design principle and framework of MyVIA, then proposes new technologies of MyVIA including User TLB, continued host physical memory and varied NIC buffer, the pipelining communication based on resource and DMA chain, and physical descriptor ring. Experimental results of performance comparisons and analysis are presented; the one-way bandwidth of MyVIA for a 4 KB message is 250 MB/s, and the lowest one-way latency is 8.46 /spl mu/s, which shows that the performance of MyVIA surpassed that of other implementations of VIA.

Extending metadata with scenarios in adaptive distributed system

The traditional metadata cannot provide enough support for the adaptive distributed systems. In this paper, we propose a metadata extension that uses the scenarios as semantic aspect, describing the expected interaction sequence based on XML Schema. With such constraints, resource waste and further damages can be avoided by preventing the illegal interactions at runtime. This extension is more valid and user-friendly than the majority of the existing proposals, avoiding the potential inconsistency. A prototype has been developed to illustrate the applicability of the extended metadata. The experiments demonstrate that the prototype introduces an average overhead ratio of 6.4% in the inter-process interactions and only 0.27% in the remote interactions. The approach presented in this paper is suitable for the adaptive distributed systems and affords more guarantee of reliability.

A Middleware Based Mobile Scientific Computing System – MobileLab

In this paper we present a scientific computing system named MobileLab which is designed for pervasive computing on the Internet. A middleware named TOM (stands for Tsinghua Operation Middleware) is designed and employed in MobileLab as a key to access the grid. Based on TOM, MobileLab can access the computational resource transparently; distribute the components to optimize the computation procedure and computation can be moved from one node to another. To use a middleware model in the architecture of scientific computing system is our new approach corresponding to the grid computing and mobile computing.

THAOP: An Aspect Oriented Programming Framework

Aspect oriented programming (AOP) is one of the hottest research areas in the development of information technology during these years. People have found many programming problems for which the traditional object oriented programming (OOP) model and component oriented programming (COP) model are not sufficient to clearly capture some of the important design decisions the program must implement. This problem is much more obvious in pervasive computing environment due to the variety of environment. AOP technology has been presented to resolve this problem. In this paper, we introduce THAOP, a lightweight and flexible aspect oriented programming framework based on TH_CORE (a.k.a. EFL) component platform, and compare THAOP to other AOP framework. And also a case study is given out to help readers have a much clearer visualize about THAOP

Extending Metadata with Scenarios in Adaptive Distributed System

The traditional metadata cannot provide enough support for the adaptive distributed systems. In this paper, we propose a metadata extension that uses the scenarios as semantic aspect, describing the expected interaction sequence based on XML schema. With such constraints, resource waste and further damages can be avoided by preventing the illegal interactions at runtime. This extension is more valid and user-friendly than the majority of the existing proposals, avoiding the potential inconsistency. A prototype has been developed to illustrate the applicability of the extended metadata. The experiments demonstrate that the prototype introduces an average overhead ratio of 6.4% in the inter-process interactions and only 0.27% in the remote interactions. The approach presented in this paper is suitable for the adaptive distributed systems and affords more guarantee of reliability

THCORE: A Parallel Computation Services Model and Runtime System

Wrapping parallel programs or parallel numerical library functions into software components and using them as computation services in service-oriented programming presents a method of delivering powerful computation capabilities of multi-processor supercomputers to the application developers who may only familiar with their desk-top or hand-held computing environment. These parallel computation services on computer clusters are used as ordinary software components on the desktop programming environment with their internal parallel or distributed characteristics hidden from the users. In order to use the parallel scientific computation applications and libraries as the software components conveniently in the development of new applications, a parallel computation service model and the runtime system that support this model on computer clusters are presented and some design and implementation issues are discussed in this article.

Increase Reliability of Pervasive Oriented Component Platform via N-Version

One of the key elements of a pervasive computing is context. Context and context inferring often cause Byzantine fault. N-version strategy is usually used to avoid Byzantine fault and improve reliability of system. We present FT-THCORE to provide pervasive oriented component specification and platform with support of N-version programming to improve the component creating and invoking mechanism

A WS-inspection based decentralized service discovery service in OGSA

Grid service discovery (GSD) is an important part of the newly proposed OGSA (open grid services architecture). This paper describes the implementation of a specific GSD service which is realized as an extended part of the OGSI technology preview. OGSI (open grid services infrastructure) has implemented the fundamental interfaces of OGSA. It provides a container registry service for the query in grid services. However, this service does not provide full flexibility for GSD, since it only supports querying among services registered to the local service container. The implemented GSD service is a decentralized one and is based on WS-inspection. It is designed as a high-level service, which is based upon the container registry service, provided in OGSI, so that direct access towards low-level data structure is avoided. The implementation details will be discussed in this paper, including the whole GSD mechanism and some more specific discovery methods used. Some furthering issues on the GSD implementation will also be discussed in this paper. Analysis on parallel and sequential searching algorithms will be given in a separate section, based on an ideal k-way tree model of the grid topology. Other possible improvements of the GSD service will be discussed as well, for example, a better use of the WSIL to extend the flexibility of the GSD service.