Today's fault management requires a sophisticated event management to condense events to meaningful fault reports. This severe practical need is addressed by event correlation which is an area of intense research in the scientific community and the industry. This paper introduces an approach for event correlation that uses a dependency graph to represent correlation knowledge. The benefit over existing approaches that are briefly classified here is that this approach is specifically well suited to instrument an existing management system for event correlation. It thereby deals with the complexity, dynamics and distribution of real--life managed systems. That is why it is considered to provide integrated event management. The basic idea is to connect the event correlator to a given management system and gain a dependency graph from it to model the functional dependencies within the managed system. The event correlator searches through the dependency graph to localize managed o...

Today's fault management is characterized by inefficient event management. The events delivered by the managed system frequently descibe symptoms of a problem instead of its cause. If a problem in the managed system occurs, e.g. a network failure or misconfigured software, the administrator often is flooded by a burst of more or less meaningless events indicating symptoms of the problem. The aim of an event correlator is to reduce the number and enrich the meaning of events shown to the administrator. Ideally, the event correlator is able to condense the received events into a single event directly indicating the problem in the managed system. The core part of any event correlation approach is a language to describe how events are (cor)related. This information is a prerequisite for feeding an event correlation engine. Many proposals have been made addressing this issue. However, as will be shown, key issues have not yet been addressed in products and research. The correlation language...

Over the past few years, network management has steadily evolved from a centralized model, where all the management processing takes place on a single network management station, to distributed models, where management is distributed over a number, potentially large, of nodes. Among distributed models, one, the weakly distributed hierarchical model, has been around for several years, whereas a flurry of new ones, based on mobile code, distributed objects or cooperative agents, have only recently emerged. Which of these techniques will eventually win ? Will several ones have to coexist ? How do they compare to each other ? In order to provide a framework to analyze these issues, this paper presents a comprehensive typology of all network management paradigms known to date, whether they have been successfully implemented already or whether they are still confined to the research community. By comparing these models with those used in another research field, enterprise management, we delineate a common trend of evolution, and attempt to predict what the future holds for network management.

The delegation of management tasks is usually accompanied with some transfer of rights. However, in many existing systems it is not possible to link this transfer with the execution of the task as the transfer usually happens in a separate independent step. Moreover, it is often not possible to control the proper use of the transferred rights. This makes it very vulnerable to abuse. In this paper we propose an approach for delegation based on mobile agents that allows control and supervision of the delegated tasks on a technical level. It will be the basis for an implementation in the Mobile Agent System Architecture (MASA).

Parallelization of image analysis tasks forms a basic key for processing huge image data in realtime. At this, suitable subtasks for parallel processing have to be extracted and mapped to components of a distributed system. Basically, this task should be done by the processing system and not by the user, as automatical parallelization allows a flexible resource management and reduces time for developing image analysis programs. This paper describes a multi-agent based system for planning and performing image analysis tasks within a distributed system. It illustrates a method for modeling image analysis tasks under the viewpoint of parallel processing and explains the special design requirements for parallelizing agents. Furthermore, we describe concepts for agent cooperation and for using the agent's ability of learning to allow long term improvement of its planning and scheduling strategies. The presented image analysis system allows an architecture-independent parallel processing of ...

Today's networks are controlled at various organisational and functional layers by human managers. What seems more suitable for the future scenarios, is a management solution based on static and/or mobile software entities, collecting network state information and which have the ability to directly invoke effective changes to switch controllers, without the interaction of a human operator. This paper focuses on the key aspects to be taken into account when distinct Network Providers need to interact in order to allocate service demands which span across distinct domains. The main motivation of our work is to define and implement a paradigm which supports automatic interaction between different network providers. The use of intelligent agents has the potential to make the interworking process much more flexible by facilitating interaction at a level above the constraints of the specific underlying network technology. By representing problems (a VPN to be set up for example) as...