RFG believes that network inventory management software, including asset configuration information and change management, is an essential component of a robust network management architecture. In addition, software tools that provide planning, design, and life cycle management for network assets should prominently appear on enterprise radar screens. IT executives should evaluate network inventory systems that provide actionable information that administrators can use to improve network management performance and help develop effective network asset control processes.

Business Imperatives:

• Typical network operations rely on physical network management that identifies element failures. However, additional key information is missing, such as device location, connectivity, and affected users. IT executives should ask the technical staff to evaluate asset management software that provides a physical and logical network view for all fault conditions, including affected network assets.
• Automated device discovery is important to catalog accurately and to update what the network actually contains. This information should be available in a central database repository with easy user access. The ability to access up-to-date network information is essential to high-reliability environments. IT executives should look for inventory management tools that report the proper level of detail as a step towards improving their network management infrastructures.
• Most networks already employ network management systems for fault detection. However, network inventory management is often missing because of network complexity and a lack of integration time. The network is an important business asset that requires ongoing tracking and management. IT executives should consider asset management tools that can integrate into the existing infrastructure, while also providing the necessary planning, design, and life cycle management capabilities.

Network complexity continues to grow as enterprises add and remove different communication technologies and multiple types of vendor equipment in response to rapidly changing e-business requirements. Software tools that identify network elements, provide logical and physical network views, track network changes, and provide accurate documentation for all network components can be collectively referred to as "knowledge-based" inventory management systems.

Several companies, including Computer Associates, Granite, NetCracker, and Visionael, have developed various types of knowledge-based inventory/asset management software that allow companies and service providers to design, document, catalog, and manage networks more effectively. As a result, tools are available to provide a better network design, improve service provisioning, manage network changes, and determine the exact location of network elements and their associated attributes. IT executives should consider network asset management as a critical piece of the network management architecture, and should ensure this capability becomes part of the network planning process.

Different network inventory vendors target different customers. Visionael provides a knowledge-based inventory and asset management solution for both enterprises and network service providers. Computer Associates, NetCracker, and Tangram are more focused on enterprise users and application service providers (ASPs), while Granite specifically targets network service providers. Regardless of the environment, network inventory management should provide concise network asset identification and configuration information for the help desk organization and network technical staff.

For example, knowing where a specific customer's router, or being able to pinpoint the location of every network device that is dependent on this router, requires a knowledge-based network management solution. This specific attribute information is valuable to help desk and technical personnel, and will help improve network availability and reliability. The ability to drill down to specific network element information should be a requirement for any inventory management system.

The first step toward effective inventory management requires a flexible discovery tool that can detect network elements and/or import asset information for all hardware, software, and infrastructure from other systems. This information should be gathered and updated regularly and stored in a central repository that allows easy access by authorized users. Since not all devices are fluent in Simple Network Management Protocol (SNMP), auto-discovery requires the use of ancillary techniques such as bar code reader capability and information transfers using a documented application programming interface (API). Some of the more popular APIs are based on open or de facto standards such as the Common Object Request Broker Architecture (CORBA), the Extensible Markup Language (XML), and Java.

Inventory information resident in a centralized database must also be coherently linked with physical and logical network information to provide a clear view of assets and locations affected when a failure occurs. For example, a router failure should be linked to specific information on the applications, users, and locations affected by such a failure.

Moreover, asset management systems must provide a consolidated physical and logical network view that provides network personnel with meaningful information in addition to an alarm indication capability. The best-of-breed products can accommodate information feeds from installed network management platforms, and can combine them with information from the asset management database. This process produces device-specific information that allows technicians to view the logical and physical sides of the network, and identifies the impact of a failure on business elements, IT assets, and users.

Some IT executives may assume that because they have sophisticated network management platforms installed, they do not require inventory management systems. However, in most cases, inventory management complements and contributes to a strong network management architecture by turning network alarms and data into useable information that identifies locations, users, and systems affected by the failure. Many vendors price and license their inventory products based on the number of objects. As a result, IT executives should ask vendors to explain their licensing scheme to determine both the initial and incremental costs for adding and managing network assets.

Likewise, a network inventory management system must be able to integrate with popular network management systems such as Aprisma Spectrum, CA Unicenter, HP OpenView, Micromuse NetCool, and Tivoli NetView that provide network fault and performance management. (See the RFG Research Note "What Happened to Integrated Network Management?," Feb. 17, 2000.) By accessing the physical network asset information, the network that relates to the active fault or alarm condition, the technical staff can immediately determine the location of the failed device or software element.

Network inventory management should also integrate into, or at least pass information to, trouble ticketing systems, such as Remedy's Action Request System. This allows device specific alarm information to be provided, as a part of the dispatch process, to a network engineer for trouble resolution, which helps reduce mean time to repair (MTTR). IT executives should explore the effort, complexity, and cost required to integrate inventory software into their existing infrastructure as part of the selection process.

Some of the potential benefits from adding inventory/asset management tools to a network management infrastructure include:

• Lower support costs resulting from proactive help desk and technical support;
• Reduced network costs from the fast identification of network elements and their attributes, including exact location and configuration information;
• Improved network availability and reliability because of a faster MTTR.

Furthermore, the asset represented by the network, whether the local area network (LAN), the metropolitan area network (MAN), or a wide area network (WAN), can be the single largest investment a company will make. Consequently, effective management of network assets makes good business sense. Inventory network management systems are also required in service provider environments to facilitate the timely identification of network assets and customers impacted by a failure or service disruption.

Although fault management platforms do a good job of detecting and identifying a network failure, the addition of inventory management software can help pinpoint where a failed device is located and the impacted users. If this information feed is integrated into help desk support systems, customers should realize improved MTTR and overall better network performance.

In summary, the following are some of the significant features of a network inventory management system that IT should thoroughly review with vendors.

• Collapsible network views that support drill-down, and that are maintained in a central database repository.
• An extensive device library providing ready-to-use models and templates for network elements, including hardware and software.
• Collaborative network design, diagramming, and documentation tools.
• A scalable, high-reliability architecture.
• Auto-discovery of SNMP-enabled devices and the ability to import inventory information from other systems and programs.
• Integration with management platforms such as Aprisma Spectrum, HP OpenView, Micromuse NetCool, and Tivoli NetView.

RFG believes that knowledge-based network inventory management is a critical network management element, and represents a best practice that IT should carefully review for implementation. IT executives will continue to face a changing network environment, and therefore, they should consider implementing software tools that can facilitate improved information flow and asset management across the network and IT infrastructures.