Electrical computers and digital processing systems: multicomput – Computer network managing
Reexamination Certificate
1998-12-09
2002-05-21
Wiley, David (Department: 2155)
Electrical computers and digital processing systems: multicomput
Computer network managing
C709S223000, C709S228000
Reexamination Certificate
active
06393472
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to the automatic aggregation of network management information in spatial, temporal and functional forms, and more particularly, to the use of an Aggregation Managed Object (AMO) in which to store attribute name-value pairs to support spatial, temporal and functional management information aggregation.
2. Description of the Art
The main challenges in providing powerful and comprehensive network management services for today's integrated networks lie in three main areas: how to provide support for heterogeneity (components of different types from different manufacturers), scalability (large numbers of network elements), and easy access to network management services (aggregation and visualization of management information).
The first challenge, heterogeneity, is being addressed through standardization efforts within organizations such as the IETF and the Network Management Forum. Today, a number of specifications are available for managing TCP/IP stacks (MIB-2 and RMON), ATM switches, etc.
There are, however, no widely established methods for dealing with the second challenge, that is with the large numbers of network elements (scalability). Managing large networks requires powerful abstractions that capture the essentials of the state of the network rather than the details. Most approaches for reducing state and event information in commercially available network management (NM) platforms are ad-hoc and usually customized for a particular management problem or network. As networks grow larger and integrate an ever increasing number of services, the existence of a scalable network management architecture becomes critical.
The first generation of network management tools to face the challenge of the large numbers of network elements (scalability), such as HP Openview, Sun Net Manager, IBM Netview, etc., follow closely the point-to-point management model. According to this model, a network management application (NM client) connects to a management agent (NM server) using one of the standard protocols for management such as SNMP or CMIP. The agent. contains information about a network element or a group of elements. A network manager retrieves or controls this information by issuing “get” and “set” operations. Especially in SNMP systems that do not support rich data types, this exchange of management information is at a very low level. As a result, all the intelligence for providing more complex NM services resides within the client (manager). First generation tools are therefore characterized by complex and expensive clients. Although these clients have the capability to maintain a hierarchical topology map and thereby provide easier navigation through a possibly large network, the manager still has to employ a low level management protocol to interact with every network element. First generation systems offer few capabilities to customize the available management services beyond the functionality offered by the underlying NM protocol such as SNMP or CMIP, or the interface provided by a vendor supplied element management system.
Second generation NM systems have tried to address some limitations of the first generation by providing better customization options and automation of routine management applications. Tools such as Tivoli's TME 10, CA's Unicenter TNG and NetExpert are targeted at the corporate IT infrastructure and offer a richer set of management services including end-to-end application management, network services management, software distribution etc.
Although the above tools simplify, to a large extent, the effort required to manage large numbers of network elements and applications, they are customized to work with specific products and network protocols. Fundamentally, the network manager has little control over what management services are presented to him/her and how information is aggregated, stored and visualized.
The third challenge then, a related area that has received much attention recently, is the one of access to network management services by the manager (aggregation and visualization of management information). Large network management systems collect a tremendous amount of information from network elements and make it available to network operators in a myriad of formats. In order for this information to convey the essence rather than the details of network state, it must be organized, summarized and simplified as much as possible. Similarly, the network manager needs mechanisms that aggregate the control of a large number of network elements into simpler interfaces.
Traditionally, network management systems have employed proprietary user interfaces to monitor and control a network state. Such systems are often customized for the specific management problem at hand and then used by a small group of appropriately trained people. This has been an acceptable solution while the only users of network management services were a small number of network operators. This situation, however, is changing rapidly: the Internet is reaching an increasing number of people and businesses every day, and broadband access is coming soon to every home. A large number of networked services are available today for businesses and consumers, ranging from simple dial-up network access to virtual networking, financial services such as online trading and banking, one step shopping, etc. The increasing complexity of online services of every form has introduced significant management requirements on both service providers and subscribers. Service providers have realized that the bundling of customer management services can be an important differentiator for their products. More customers today require the ability to observe the operational state of their service in real-time, collect statistics on service usage, customize parameters of the service, order additional service or perform proactive management tasks in anticipation of efforts. By delegating some aspects of managing a service to customers, operators can cut down on their customer care costs while providing competitive and cost effective services.
The increasing availability of management services has motivated many researchers to rethink the way these services are provided to consumers. The continuation of the existing status quo which calls for customized and complex user interfaces to service management functions receives increasing resistance from businesses and consumers that favor portable, lightweight, standards-based solutions that need the minimum amount of configuration and are simple to use. Many researchers have proposed to use the World-Wide Web (WWW) to provide access to management services. The Web offers the widest possible installed base of compatible clients (every networked computer is now equipped with a Web browser) and a portable execution environment based on Java that allows Web clients to access arbitrarily complex information services by downloading the appropriate Java applets.
Access to the management services has thus been provided using the World-Wide Web and Java, the most widely available tools today for remote information access. A resulting software platform for such access was named Marvel (for Management AggRegation and Visualization Environment) which is detailed in the reference “MARVEL: A Toolkit for building Scalable Web-based Management Services”, and which is hereby incorporated by reference. Marvel is a software environment that allows the network manager to define how management information collected from network elements is aggregated into more useful abstractions and finally presented to the manager. Marvel thus provides scalable solutions for systems management for small businesses and large enterprises, network management services for network operators, and customer network management services for businesses and consumers.
The MARVEL architecture consists of lightweight clients and a hierarchy of Management Aggregation and Visualization Servers (MAVS). The minimum requiremen
Anerousis Nikolaos
Biliris Alexandros
Jagadish Hosagrahar Visvesvaraya
Dinh Khanh Quang
Wiley David
LandOfFree
Automatic aggregation of network management information in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Automatic aggregation of network management information in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Automatic aggregation of network management information in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2894458