TUTORIALS

Program at-a-glance

Monday Morning T1: Management Platforms and Service Management T2: Management of Service Level Agreements T3: Event Correlation Technologies Management T4: Java Management Extensions (JMX)
Monday Afternoon T5: Internet Management Protocols T6: Policy Based Network and Security Management T7: Designing for Effective Event Management  
Friday Morning T8: Architectures, Control, and Management of IP-over-Optical Networks T9: Common Information Management (CIM) T10: Active Network Techniques in Network T11: Shattered Security: A Clear View Through Glass Fortresses
Friday Afternoon T12: Virtual Private Networks T13: Directory Enabled Networks (DEN) T14: Jini-based Management of Networks and Distributed Systems  

T1: Management Platforms and Service Management
Joseph Ghetie, Telcordia Technologies, USA

This tutorial analyzes the state of management platforms and their evolution from network management to system management and service management. Service management concepts, criteria, and metrics associated with Quality of Services and Service Level Agreements are also discussed. The tutorial explores the capabilities of the new generation of management platforms in providing solutions and applications focused on service management of enterprise-wide and telecommunications networks.

T2: Management of Service Level Agreements
Lundy Lewis, Aprisma Management Technologies, USA

A current trend among enterprises, service providers, and Telcos is the development of service level agreements. SLAs are contracts between providers and consumers -- replete with punishment and reward clauses relative to the degree to which agreed services are met. However, vendor support of SLA management is fragmented. Some vendors promote statistics reporting as the essence of SLA management, while other vendors promote application monitoring, service development, business process re-engineering, supplier/consumer negotiation, or contract development as the essence of SLA management. In this sort of situation, a general definitive framework that subsumes piecemeal approaches does much good. In this tutorial we provide a framework that serves as a baseline against which one may situate, evaluate, and execute SLA programs. We provide (i) definitions that lead to the broader concept of Service Level Management (SLM), (ii) a general SLM process, (iii) an SLM architecture, (iv) research challenges, (v) current tools and applications, and (iv) case studies.

T3: Event Correlation Technologies Applied to Advanced Network and Service Management
Gabriel Jacobson, Verizon Laboratories, USA

This tutorial is targeted to a wide audience of network management R&D personnel, network operators, and vendors. The tutorial discusses the issues and solutions of next generation event correlation systems, their applications and enabling technologies. Event correlation is a widely accepted technology for managing the complexity of modern telecommunication and data networks. It has been mostly used for network fault detection and root cause analysis; however, the most recent applications have broadened the scope of event correlation including the network performance and service quality management. The first part of the tutorial explains the basic concepts of real-time event correlation, describes various approaches to event correlation, including rule-, case-, and model-based reasoning, finite state machines, and binary coding methods. We will also discuss the architecture of distributed correlation services. The second part reviews existing products of event correlation and describes the process of development and deployment of event correlation applications. The third part of the tutorial is devoted to the new research issues and future trends in the application of event correlation systems.

T4: Java Management Extensions (JMX)
Swee Lim, Sun Microsystems, USA

As Java moves from "cool" to "critical," the need to manage Java based devices and applications has become the latest challenge for the enterprise. Java Management Extensions (JMX) provides the management instrumentation APIs that are needed to create manageable applications of any class while maintaining a short learning curve, low programming burden (as little as a few lines of code!), and management system vendor independence. JMX has been targeted to support embedded to enterprise Java resources. In this tutorial we will briefly introduce application management concepts and identify which management needs JMX is intended to satisfy; provide a technical overview of JMX technologies, its compliance levels, its components, and its future; review the JMX APIs and class structure..

T5: Internet Management Protocols
Aiko Pras, University of Twente, The Netherlands

This tutorial presents the state-of-the-art of Internet management protocols, as being defined within the IETF. It discusses the history of such protocols, and reviews the developments with respect to the Structure of Management Information (SMI), Management Information Bases (MIBs) as well as the Simple Network Management Protocol (SNMP). In particular it discusses the differences between SMIv1 and SMIv2, the MIBs that have been derived from the MIB-II, and the development from SNMPv1 via SNMPv2 to SNMPv3. Other topics that will be addressed are extensible agent technology (AgentX) and distributed management (DisMan). The tutorial concludes with a discussion of recent developments within the IETF and IRTF management groups.

T6: Policy Based Network and Security Management
Morris Sloman, Emil Lupu, Imperial College, UK

There is considerable interest in how to specify policy for management of multi-service networks, role based access control and trust for e-commerce. In management systems, interpreting policy rather than coding it into the management components provides a means of implementing dynamically adaptive bandwidth allocation and Quality of Service (QoS) management. Policies can be changed dynamically in order to change the management strategy and behavior of the management system. Role Based Access Control (RBAC) provides a flexible means of grouping permissions relating to positions in the organization so that people can be assigned to roles without changing policy. This half-day tutorial will explain what is policy, why it is needed and will give an overview of the policy work going on in both the network management and security communities. This will include the RBAC, model, Trust Policy Language (from IBM), and the Policy Definition Language from Lucent. The tutorial will present in more detail the Ponder Policy Framework (see http://www-dse.doc.ic.ac.uk/policies) which provides a declarative, object-oriented language for specifying both security and management policies. Ponder, supports a rich set of constructs for grouping policies in roles, relationships and management structures. We will cover the issues relating to refining high-level enterprise goals into implementable policies; detecting and resolving conflicts and how to specify semantic meta-policies or constraints on the permitted policies within the system to cater for conflicts of duties etc., as well as issues relating to implementation and deployment of policies in a distributed environment.

T7: Designing for Effective Event Management
David Thoenen, IBM Corporation, USA

It is commonly accepted that the impact on the corporate bottom-line of IT business application services is directly proportional to the up-front investment in analysis of system requirements and development of detailed implementation specifications. Experience now teaches that this is also true for the implementation of event management systems. This tutorial will present a step-by-step process for assessment of event processing and correlation requirements and development of detailed specifications for their implementation. The tutorial will draw upon examples collected over the course of over fifty successful design projects. Steps within this design process to be addressed include: Defining scope of the design; why and how to focus upon specific business services. Developing enterprise policy for event processing; why and how to use business principles as a foundation for technical decisions. Researching and documenting event repertoires; why the need for comprehensive event analysis (and how to do it quickly). Applying policy when making event filtering decisions; how to view event filtering as a business exercise and achieve dramatic results in event volume reduction. Utilizing systematic methodology for event correlation analysis; why correlate, how to determine which events require correlation and how to decipher their correlation requirements.

T8: Architectures, Control, and Management of IP-over-Optical Networks
Chien-Chung Shen, University of Delaware, USA

Driven by the demand for network bandwidth (customer pull) and the advances in WDM technology (technology push), the next generation Internet is expected to be an IP-based optical WDM network, where the integration of gigabit/terabit IP routers with WDM switching and transmission systems creates optimized transport networks to support Internet's phenomenal growth. The shift from a static point-to-point architecture towards a dynamic IP over re-configurable WDM architecture raises new issues and demands new techniques from the network control and management perspective, The tutorial will begin with an introduction to IP and optical networking technologies (gigabit/terabit IP routers and optical cross-connects), and describe the interconnection models of IP over re-configurable WDM network architecture. The tutorial will then discuss control and management issues of IP over re-configurable WDM networks and describe multi-protocol lambda switching for optical cross-connect, configuration management and routing, fault management and restoration, and performance management and traffic engineering. Recent research on IP over Switched WDM (optical burst/label switching) will also be described.

T9: Common Information Management (CIM)
Andrea Westerinen, Cisco Systems, USA

The Common Information Model (CIM) from the Distributed Management Task Force (DMTF) has undertaken an ambitious task - to describe the "managed environment," from computer systems and their devices, to software, users, networks and policy. This tutorial overviews the design of CIM, its importance to WBEM and DEN, and how to get started using CIM. Each of the component models (such as Core, System, User and Network) are discussed at a high level. Their design goals and scope are summarized.

T10: Active Network Techniques in Network Management
Danny Raz, Technion, Israel, and Lucent Technologies, USA

"Active networks" denotes a framework where network elements, primarily routers and switches, are programmable. Programs that are injected into the network are executed by the network elements to achieve higher flexibility for networking functions, and to present new capabilities for higher layer functions by allowing data fusion in the network layer. In this tutorial, the technology developed recently by the active network research community will be surveyed. The tutorial will focus on the various ways this technology can be used both to develop better NM tools and to support fast development and deployment of value-added services in the network. Relevant standards efforts for active networks and programming interfaces (such as ANEP, and IEEE P1520) will be summarized.

T11: Shattered Security: A Clear View Through Glass Fortresses
Michael S. Greenberg, Sandstorm Enterprises and Theophany Holding, USA

If there was an Information War, and your network was part of the battlefield, would you know? The answer is not encouraging. Based on the old information fortress model, the current computer security paradigm is insufficient to secure computers in the modern distributed and networked computing environment. The current security situation is in crisis. Developing technologies -- such as mobile agents, objects, and embedded macros -- further strain automated security policies while faster and better connectivity makes detection of security breaches more complex. This tutorial will cover the fundamentals of the modern security paradigm and how it fails in the current computing environment. We will describe the existing computing environment, and the threats and risks associated with it. We will survey the fundamental building blocks of computer security, and how they are commonly integrated to protect machines and networks. We will then go into depth on techniques used to attempt to secure mobile agents and the hosts they visit to expose the strengths and weaknesses of the current security paradigm. We will investigate the reasons the current paradigm is insufficient, analyze malicious programs and strategies, and survey proposed solutions. To conclude, we will discuss the need for a new paradigm in computer security and how security and network management techniques may be used together to attempt to improve the security situation.

T12: Virtual Private Networks: Architectures, Design, Planning and Management
Raouf Boutaba, University of Waterloo, Canada

This tutorial provides a practical understanding of virtual private networks (VPNs) and a foundation for evaluating them from the perspectives of business benefits as well as technical requirements. Examples from real-world experiences provide an insightful view of the advantages of this technology and the pitfalls to avoid in choosing or designing a VPN. The tutorial particularly emphasises the technical aspects related to the planning, implementation, operation and administration of VPNs.

T13: Directory Enabled Networks (DEN)
John Strassner, Cisco Systems, USA

This tutorial will provide a brief overview of directories, and focus on how to map an information model into a form that can be stored in a directory and manipulated using the LDAPv3 access protocol. It will then briefly review the DEN policy model and explain how that model can be mapped into a directory. It will then provide a detailed example of how the model can be used in practice. The tutorial will conclude with the overview of how policy system can be implemented using CIM, DEN and Ponder Policy Framework.

T14: Jini-Based Management of Networks and Distributed Systems
Gerd Aschemann and Peer Hasselmeyer, TU Darmstadt, Germany

Jini, a Java-based network infrastructure from Sun Microsystems, seems to be very promising - if used appropriately - to become a new enabling technology for the management of networks, distributed systems, and distributed applications. In the future there will be a huge market for small and medium sized management solutions which need to integrate several technologies and existing products. Much more automation of management tasks, particularly integration of new entities (devices and services), and configuration of managed systems will be required. Jini enhances such integration and automation and has the ability to replace the monolithic management platforms and huge management frameworks of the past by a well-suited collection of small and adequate management services (or at least to split them up accordingly). Even in the field of large-sized management solutions, a new Jini-based standard has recently been released: the Federated Management Architecture (FMA). This tutorial will introduce Jini and related technologies, e.g., Java Spaces and FMA, and investigate Jini from the viewpoint of management, i.e., show how it could enhance typical management tasks on the one hand, and on the other hand, show how federations of Jini-enabled devices and services could be managed.