Performance and Fault Management: A Practical Guide to Effectively Managing Cisco Network Devices (Cisco Press Core Series) - page 42

Summary

The previous sections covered many kinds of information you will want to include in your knowledge base. This summary brings this information together.

Taken together, all the information that you desire to store in your knowledge base will define its schema. Collecting this information as you define your needs for your network will allow you to define a custom schema for your knowledge base. Table 3-1 compiles the information from the preceding sections into an example schema.

Table 3-1. Network Knowledge Base Schema

Choose a method of implementing your knowledge base that starts out simply, yet can expand as you expand the amount and type of network management activities. Keep things simple and only implement the items that are more difficult if they are truly required in your network.

The bottom line for all this information is use it or lose it. If you are not actively using this information, it will quickly become out-of-date and useless. So, rely on this information and you will keep it accurate.

Now that you have a storage vehicle for information about your network, you can use it. The next chapter discusses performance management of your network, and will rely on the knowledge base to determine what to collect and to store the collected data.

Chapter 4. Performance Measurement and Reporting

As companies have migrated from centralized mainframe-based networks with strong performance- related tools to distributed networks, they have discovered the increased complexity of trying to monitor and analyze performance data gathered from varying vendors ' devices.

For example, mainframe availability is a key performance measurement for traditional SNA networks. IT groups measure the availability of their mainframe, its applications, and its network connections and controllers. They then roll these numbers into a single availability number. This is the coveted "five nines," or 99.999 percent, which some IT groups try to achieve.

Upon moving to a distributed networking model, data and applications are distributed across multiple file servers and the network is segmented into multitudes of routers, switches, and concentrators . The determination, collection, and reporting of performance-related data have become considerably more difficult. The capability to roll up performance analysis into reports that can be consumed by upper management is equally difficult.

Fortunately, as methods and equipment for distributed networks mature, methodologies and tools that collect and report on performance data have become more accessible.

This chapter covers the following aspects of performance measurement and reporting:

• What performance management is

• Performance data collection

• Performance data reporting

• Measuring network performance