With the fast spreading of Grids in both research and industry, there is also an increasing demand for an adequate management of both Grid infrastructures and Virtual Organizations. Some of the new management challenges associated with Grids stem from their very nature: resource providers are usually independent, autonomous organizations exhibiting different legal, financial, and interest backgrounds; Grid resources are typically heterogeneous with unknown dependability patterns; Virtual Organizations need to be created, modified, destroyed dynamically; workflows need to cross organizational boundaries; operational processes must be enabled to adapt themselves dynamically to current environmental and situational changes.

The MNM Team focuses their Grid research activities on these questions. In particular the team is investigating how and to what extent the necessary flexibility of integrated Grid management may be automated. Research goals are thus the provisioning of a framework for managing dynamic Virtual Organizations, to understand the Grid co-management problem (resources can belong to several management domains), the development of adequate concepts for the self-management of resources and services, monitoring and accounting within the context of Grid environments, and the provisioning of a management specific Grid testbed.

Cooperations:

The MNM Team has joined several national and international cooperations and projects related to Grid research. Examples are: the Distributed European Infrastructure for Supercomputing Applications (DEISA), a consortium of leading national supercomputing centres; the D-Grid initiative for the promotion of a Grid-based e-science framework in Germany; the ByGRID, a project for sharing the supercomputer resources of the Bavarian universities; the Atlas Data Challenge/NorduGrid, a Grid project for wide area computing and data handling associated with CERN's Large Hadron Collider Grid; the Uniform Interface to Computing Resources (UNICORE/UNICOREplus) Forum.

In addition, the MNM Team contributes to the Global Grid Forum (GGF) and participates in GridLab and AccessGrid activities.

This is a Network of Excellence project supported by the European Commission Information Society Technologies 6th Framework Program. The network brings together 13 of the best research teams across Europe in the discipline of network and service management.

It addresses the scalability, dynamics, security and automation challenges that emerge towards the management plane of the future Internet.

Its main objective is to extend the European leadership and expertise in the area of the network and service management plane, critical to every network.

Research integration is achieved within the project through both the joined establishment of a common vision of major challenges in the discipline, complemented with specific targeted joint research actions addressing the following management themes:

• Management Foundations
• Management Technologies
• Management Applications

The Wissenschaftsnetz (X-WiN, Scientific Research Network) is the national part of the German research network and is therefore the heart of "the Internet for science in Germany". It spans an infrastructure among more than 40 core network routers which are under the responsibility of the DFN-Verein.

To give all institutions (users) that are connected to the X-WiN an insight into the general state of the network as well as into the exact data for their own link the DFN-Verein provides a Customer Network Management (CNM) system which is run at the Leibniz Supercomputing Center. All users get access to this system automatically.

For the IP service "DFNInternet" the CNM contains the following applications for the users:

• Data Volume: With the help of this application the users can track how much IP traffic they have send or received via their X-WiN link. This application is completely implemented and is available for all users.
• Topology: With the support of this application the users can get current information about the status of the X-WiN. In addition, it is possible to get historic data. This application is implemented and is available since April 2004.

Since the start of the EU-funded project GN2 JRA1 in September 2004, the CNM tool has been evolved for its use in the European research backbone network Geant2 and connected national research and education networks. The topology application shows the network topologies of these networks and shows utilization as well as delay, jitter, and packet loss data.

Cooperation partners: Deutsches Forschungsnetz (DFN), DANTE

Most existing security models do not consider cases of temporary, single task delegations where a delegate receives a task under special exceptional circumstances, for which the delegate is not fully authorized. This scenario has several implications on security: on the one hand the delegate needs all necessary rights to execute the task, on the other hand the delegating is still responsible for all actions taken in the context of the task. This project investigates different ways how to solve these problems securely.

Cooperation partners: Deutsche Forschungsgemeinschaft (DFG), Technische Universiät München (TUM)

Modern large-scale research projects such as the Large Hadron Collider (LHC) at CERN require a tight international cooperation and have exceeding network bandwidth demands. Dedicated optical multi gigabit End-to-End (E2E) Links is a new service within the pan-European research network Géant2 which offer a possible way to meet these requirements. These links generally are realized with layer 1 and 2 technologies and cross multiple networks operated by different organisational domains. The federated organisation of the operation and the usage of heterogeneous hardware applying a variety of information access and management techniques require a new approach for the management of these multi-domain E2E Links.

Commencing end of 2005 a general concept for management of E2E Links was elaborated and for this concept required E2E Monitoring System (E2Emon) was developed. Meanwhile E2Emon is deployed in the production environment of the E2E Coordination Unit (U2ECU) which coordinates multi-domain activities on E2E Links.

Cooperation partners: Géant2, DANTE, Deutsches Forschungsnetz (DFN), Communication et Systèmes (C-S)

• Trend analysis for Virtualization Solutions
• Virtualization solutions in embedded systems
• Management of virtualization solutions
• Security of virtualization solutions

Cooperation partners: Siemens Corporate Technology

At the same time as devices in the information and communication technology branch are getting more and more complex, more and more self-management functions are intgrated into hard- and software, hiding much of the complexity from the user. In this project, we investigate the cooperation of external control and self-management.

One goal of the project is to establish a classification of systems with self-management functions. Therefore, metrics have to be defined for the various criteria. The work shall result in an evaluation sheet, that enables the comparison of such systems. Engineering disciplines such as automotive engineering, aviation and space technology shall be investigated for concepts, that are promising to apply to IT systems management.

The knowledge, gained in the first part shall then be applied to a practical problem. The chosen scenario is the automation of update management in Smart Home environments. Update management is therefore modelled as a generic process independent of application domain. Policies allow to control the update management process.

The mobile telecommunication sector has evolved rapidly during the last decade, spawning new technology, standards and constantly evolving business models and services. While telephony is still by far the most important service, networks providing higher bandwidth to the end user allow the deployment of sophisticated value added services.

To allow a seamless operation of these services, providers need to implement inter-domain service and device management. The advent of (hitherto mostly desktop-based) wireless technologies such as Wireless LAN (aka WiFi) and Bluetooth add to the heterogeneity of the networks and thus increase the management challenge.

The goal of this project is to analyse the application of policy-based management techniques to 4th Generation mobile communications infrastructures and devices. This includes the development of distribution concepts for policy-based management architectures as well as an evaluation of suitable policy languages.

KODI (Kontextsensitive Dienste in globalen UMTS Diensten) – a German abbreviation for Context-aware Services – is a project whose goal is to introduce the concepts of context-awareness to services in the UMTS environment.

Context-awareness describes the ability of a service to adapt to the situation of the user. As a result, interaction with the user is optimized and the acceptance of mobile services is likely to increase.

UMTS stands for Universal Mobile Telecommunications System and denotes the Third Generation of cellular telecommunication networks. It evolves from today’s widely deployed GSM (Global System for Mobile Communications).

The KODI-Project is a joint effort of Siemens AG Corporate Technology, Apollis interactive AG and the Ludwig-Maximilian-University of Munich and is partially funded by the State of Bavaria (Bayerische Staatskanzlei).

Despite some disillusionment after the burst of the dot-com bubble regarding the short-term success of ASPs, the case for the ASP business model remains persuasive. Compared to traditional, in-house IT service provisioning, it offers economies of scale and scope. It still is the most touchable part of the IT as a utility-vision.

However the ASP model poses several IT management challenges, especially pertaining to Service Level Management. This project seeks to identify and investigate the research topics, on a organizational as well as a technical level, that arise from the need of managing the Quality of Service (QoS) across inter-organizational service-chains.