The On-line Monitoring Interface Specification (OMIS) aims at defining an open interface for connecting on-line software development tools to parallel programs running in a distributed environment. Interactive tools like debuggers and performance analyzers and automatic tools like load balancers are typical representatives of the considered class of tools. The current situation is characterized by the fact that tools either follow the off-line paradigm by only having access to trace data and not to the running program or else they are on-line oriented but suffer from the following deficiencies: they do not support interoperability in the sense that different tools can be used simultaneously --- not even tools from the same developer. Furthermore, no uniform environment exists where the same tools can be used for parallel programs running on different target architectures. A reason for this situation can be found in a lack of systematic development of monitoring systems, i.e. systems wh...

THE TOOL-SET for PVM will comprisea set of integrated tools which can either be used individually or in concert. THE TOOL-SET will be composedof a debugger, a performance analyzer, a visualizer, a deterministic execution controller, a load balancer including a checkpoint generator, anda parallel file system. All tools will be available under the GNU General Public License Agreement. First versions of the tool environment will be released in spring 1996. 1 Introduction PVM has become a de-facto standard for writing parallel applications based on the message passing paradigm. It is also being used as a platform for developing tools for parallel programming. Although some of them are rather sophisticated, currently only few are really usable for application programmers. One reason for this fact is that many tools are pure research prototypes providing only rudimentary, clumsy interfaces. A second reason is that they cover only a single aspect of parallel program development and don't sup...

In order to develop and maintain parallel and distributed software systems, tools that observe and manipulate the run-time behavior of these systems are indispensable. Often it would even be preferable use multiple tools at the same time in order to have enhanced functionality. Today, tools allowing such a simultaneous use only exist in the form of monolithic integrated environments produced by a single vendor. As soon as we want to combine independent tools we are faced with severe problems. The paper presents the features included in the OMIS compliant monitoring system OCM that help to solve these problems. We derive the necessary features by looking at an example scenario, where several tools are used together to provide an efficient debugging environment, allowing the user to reset the debugged program to any previously executed state. This scenario also shows that tool interoperability is not only a feature by itself, but also leads to an added value for the user. 1 Introductio...

. Debuggers are critical tools for the software development. The design and implementation of a source-level debugging system that enables the HPF programmer to observe the behavior of the program at the level at which the program has been developed presents unique challenges. The main requirement put on an HPF debugger is to observe and control the state of many processors, to summarize and present distributed information in a concise and clear way in terms of the source program to the user. To be practical, the debugger has to support interactive source-level debugging of large-scale applications on large machines. In this paper we define design goals for the HPF debuggers and present an architecture of an advanced HPF debugging system DeHiFo, which addresses several challenges involved and provides significant contributions into the existing debugging technology. The HPF debugger is a rather complex system. Its development requires a systematic cooperation between seve...

. On-line tools for parallel and distributed programs require a facility to observe and possibly manipulate the programs' run-time behavior, a so called monitoring system. Currently, most tools use proprietary monitoring techniques that are incompatible to each other and usually apply only to specific target plattforms. The On-line Monitoring Interface Specification (OMIS) is the first specification of a universal interface between different tools and a monitoring system, thus enabling interoperable, portable and uniform tool environments. The paper gives an introduction into the basic concepts of OMIS and presents the design and implementation of an OMIS compliant monitoring system (OCM). 1 Introduction The development and maintenance of parallel programs is inherently more expensive than that of sequential programs. This is partly due to the lack of widely available, powerful tools supporting test and production phases of those programs, e.g. performance analyzers, debugge...

Nondeterminacy implies two unpleasant properties for testing and debugging parallel programs: successive executions of the same program with the same input values often do not show identical behaviour and watching the program influences the execution. It is therefore not always possible to test and debug parallel programs like sequential ones. This article presents an approach how to make test, regression test and cyclic debugging as easy as in the sequential case. To meet this goal we provide the user with the ability to specify not only input data for tests, but also the sequence of interprocess communications. This also guarantees reproducibility for debugging and regression testing. 1 Introduction Massively parallel systems gain more and more success in commercial markets and try to deal with the Grand Challenges [GC92]. But despite all success it is still hard to exploit the available computing power of these systems. Besides the fact that design and coding of parallel pro...

The copyright for this contribution is held by

. Source level debuggers for parallel PVM or MPI programs currently offer good support for debugging multiple processes, however, they still lack adequate mechanisms for debugging message passing errors. In this paper, we present a new concept called message breakpoints, which allows to follow the information flow between processes. We also show how these breakpoints can be implemented very efficiently by using invisible message tags and will demonstrate other applications of this technique in the context of distributed event detection. 1 Introduction One of the major tasks during the process of software debugging is to follow a program's data and information flow in order to find out why it behaves the way it does. This is usually done by stopping the program at interesting points and executing it step by step while watching the changes in program state, e.g. the values of variables. This process is supported by source level debuggers [12], which now are also available for parallel ...

Debuggers are critical tools for the software development. The design and implementation of a source-level debugging system that enables the HPF programmer to observe the behavior of the program at the level at which the program has been developed presents unique challenges. The main requirement put on an HPF debugger is to observe and control the state of many processors, to summarize and present distributed information in a concise and clear way in terms of the source program to the user. To be practical, the debugger has to support interactive source-level debugging of large-scale applications on large machines. In this paper we define design goals for HPF debuggers and present an architecture of an advanced HPF debugging system DeHiFo, which addresses several challenges involved and provides significant contributions to the existing debugging technology. The HPF debugger is a rather complex system. Its development requires a systematic cooperation between several partne...

The Tool-set for PVM will comprise a set of integrated tools which can either be used individually or in concert. Tools will be subdivided into interactive tools for investigation and manipulation of the program state and automatic tools. Both types of tools will use online data, trace data, and checkpoint information to perform their task. The Tool-set will be composed of a debugger, a performance analyzer, a visualizer, a deterministic execution controller, a load balancer including a checkpoint generator, and a parallel file system. All tools will be available under the GNU General Public License Agreement. First versions of the tool environment will be released in spring 1996; some of its components will be available earlier. 1 Introduction The PVM programming library has become an increasingly popular de-facto standard for writing explicitly parallel programs based on the message passing paradigm. A large variety of parallel applications, e.g. multigrid solvers, image p...