Proactive Network Management

As stated earlier, no network is completely fault-tolerant. Problems inevitably arise, interfaces go down, and bottlenecks occur. During times like these, a network management solution is often a network administrator’s best friend, allowing easy identification and resolution of network outages.

Network management is more than troubleshooting problem spots on a network, however. Today’s network management solutions provide a host of tools that can perform detailed analysis on a network’s performance and status, providing valuable information to network managers. In addition, these tools can identify potential trouble spots (faulty interfaces, data bottlenecks, etc.) before they threaten network up time.

To utilize these tools in a proactive manner, network managers should initiate the following actions to better gauge the performance of their systems:

Set goals Considering the amount and types of equipment on your network, as well as the configurations of each, it is wise to set a network performance goal based on a number of different metrics. Metrics can include link reliability, route optimization, effective use of bandwidth, router CPU usage, or any number of measurable variables.

Generate reports Reports are the lifeblood of any well-maintained network management scheme. Using data gathered from the network management solution, network managers can compose reports that provide as detailed a summary as needed of network performance. These reports should be weighed against the stated performance goals, and the collection of information used in network reports should certainly reflect the metrics that define the design goals. In other words, knowing what to look for can eliminate sifting through mounds of unneeded data.

Identify problems Perhaps the most vital feature of network management is the ability to identify not only severe problems (i.e., network outages), but also potential bottlenecks and/or irregularities. In this way, performance reports generated from data analysis can be used for future upgrades and improvements in a network before they become a problem.

Respond to change A network is an ever-changing entity, with components being removed and added often. A seemingly innocent change on a large internetwork can create a ripple effect that can dramatically alter a network’s performance. Network management trend analysis provides summaries of network status over set periods of time. As the network grows, trends may require that you rethink the report metrics. For example, the manager of a small network using only static IP routes might consider implementing a routing protocol such as RIP or IGRP as the network grows and adds a number of links.