Middleware For Distributed Systems Research Articles

A Distributed System for The Management of Fine-grained Provenance

Existing provenance systems operate at a single layer of abstraction (workflow/process/OS) at which they record and store provenance. However, the provenance captured from different layers provides the highest benefit when integrated through a unified provenance framework. To build such a framework, a comprehensive provenance model able to represent the provenance of data objects with various semantics and granularity is the first step. In this paper, the authors propose a provenance model able to represent the provenance of any data object captured at any abstraction layer and present an abstract schema of the model. The expressive nature of the model enables a wide range of provenance queries. The authors also illustrate the utility of their model in real world data processing systems. In the paper, they also introduce a data provenance distributed middleware system composed of several different components and services that capture provenance according to their model and securely stores it in a central repository. As part of our middleware, the authors present a thin stackable file system, called FiPS, for capturing local provenance in a portable manner. FiPS is able to capture provenance at various degrees of granularity, transform provenance records into secure information, and direct the resulting provenance data to various persistent storage systems.

Research on the Message-Oriented Middleware for Wireless Sensor Networks

The Propagation of information in wireless sensor networks (WSN) is critical to the communication. It is necessary and important to design a suitable communication primitive. Among these communication primitives, one distinguished primitive is Publish/Subscribe Communication Generics. This paper introduces current middleware of wireless sensor networks; analyze layer structure and problems of middleware. Then we design a message-oriented middleware system named WMOS (Wireless Message-Oriented System) which is developed on TinyOS. WMOS has three features mainly: first, WMOS is a distributed middleware system which is well-adapted to scalable system. WMOS utilizes conception of fully distributed and modular to make it appropriate for practical implementations. Second, we supply self-adapt multilevel QoS service to save power effectively. Third, we introduce topic/content double mode based on XML matched in WSN application. It guarantees effective query information than other middleware.

Architecture and performance of RPC-based distributed vibration monitoring and diagnostic system

The paper presents a proposal of the architecture of the distributed vibration monitoring and diagnostic system. In larger distributed systems middleware technologies are used for the data exchange. The paper presents advantages and disadvantages of the three popular technologies: CORBA, RPC and DCOM. The author proposes RPC as the preferred solution for the distributed monitoring and diagnostics system. The biggest advantages of RPC are: low overhead, platform independence, relatively easy usage and high popularity. Next, the architecture of the distributed vibration monitoring and diagnostic system is proposed and described. The system based on the proposed architecture was developed during the research project. The important feature of the architecture is a general approach to data processing and a possibility of easy extension of its functionality. The last part of the paper presents performance measurements of the system. Time required for data processing and data exchange by RPC were measured and compared. The tests were performed in six different configurations. The test results are presented and discussed. The tests showed that RPC is a very efficient and scalable middleware technology which improves design and parameters of the distributed vibration monitoring system.

Towards the measurement of tuple space performance

Many applications rely upon a tuple space within distributed system middleware to provide loosely coupled communication and service coordination. This paper describes an approach for measuring the throughput and response time of a tuple space when it handles concurrent local space interactions. Furthermore, it discusses a technique that populates a tuple space with tuples before the execution of a benchmark in order to age the tuple space and provide a worst-case measurement of space performance. We apply the tuple space benchmarking and aging methods to the measurement of the performance of a JavaSpace, a current example of a tuple space that integrates with the Jini network technology. The experiment results indicate that: (i) the JavaSpace exhibits limited scalability as the number of concurrent interactions from local space clients increases, (ii) the aging technique can operate with acceptable time overhead, and (iii) the aging technique does ensure that the results from benchmarking capture the worst-case performance of a tuple space.

The more things change . . .

Developing any nontrivial middleware-based distributed system is hard, regardless of the middleware technology. Although technology's continuous march should make systems easier to develop, somehow the details always remain challenging. New technologies only seem to make simple things simpler; they don't help with the more complex things. Horizontal concerns, such as security, transactions, fault tolerance, load balancing, and enterprise management, never seem to get easier, regardless of what middleware development platform you choose. Dealing with versioning and change management in a deployed middleware application can be complicated and costly. A good example of versioning coming into play for distributed middleware systems is the interface between any two applications or components. Interface versioning issues apply whether the middleware explicitly employs an interface definition language (IDL), as in distributed-object systems like Corba and Microsoft COM, or whether "interfaces" really are represented as exchanged documents, as in many messaging systems.

Exploiting high-level coherence information to optimize distributed shared state

InterWeave is a distributed middleware system that supports the sharing of strongly typed, pointer-rich data structures across a wide variety of hardware architectures, operating systems, and programming languages. As a complement to RPC/RMI, InterWeave facilitates the rapid development of maintainable code by allowing processes to access shared data using ordinary reads and writes.Internally, InterWeave employs a variety of aggressive optimizations to obtain significant performance improvements with minimal programmer effort. In this paper, we focus on application-specific optimizations that exploit dynamic high-level information about an application's spatial data access patterns and the stringency of its coherence requirements. Using applications drawn from computer vision, datamining, and web proxy caching, we illustrate the specification of coherence requirements based on the (temporal) concept of "recent enough" to use, and introduce two (spatial) notions of view s, which allow a program to limit coherence management to the portion of a data structure actively in use. Experiments with these applications show that InterWeave can reduce their communication traffic by up to one order of magnitude with minimum effort on the part of the application programmer.