CrossNodes Briefing: Network Management Software

Most network administrators want to prevent problems before they occur. Unfortunately, the nature of networks dictates a reactive approach. The problem of the hour forces administrators to find an immediate solution and leaves them wishing that they could anticipate these moments of crisis. Network management utilities and software can help.

Network management programs vary greatly in scope and power. Some are limited to monitoring traffic patterns or server performance. Others offer comprehensive services that measure the health of the entire network. Although all take time to install and operate, most users agree that they are a necessary diagnostic tool for maintaining and optimizing a network. Specific functions performed by the software applications can include:

  • Network mapping — provides a graphical look at the network configuration. This is especially helpful for dispersed networks with remote servers.
  • Remote notification — sends an alarm automatically to the administrator using a beeper, a pager, or an e-mail address. Network administrators potentially can get an alarm and solve a problem before a user notices.
  • Load balancing — offers utilities that permit network administrators to prioritize traffic and balance work demands across several servers. These utilities can work automatically or may require user intervention.
  • Historical database — maintains a log of network usage and performance statistics that allow network administrator to identify changes in the performance of network components. These changes may indicate that the device is failing, and the administrator can take steps to replace or repair the device before it disrupts the network.
  • Current and historical errors — generates a database of problems. This can help the network administrator diagnose problems, locate and correct bottlenecks, and improve inefficient network configurations.
  • Inventory management — provides a current listing of software and hardware running on the network. Administrators can use this utility to better control the network and ensure that workstations and servers meet the minimum requirements of the network.
  • Reporting — generates reports detailing network operation. Administrators responsible for complex networks frequently prefer graphical reports that help analyze network performance and problems.

The Smart Approach
Network management software depends on the components installed in the network. Many “smart” bridges, switches, routers, hubs, and workstations support Simple Network Management Protocol (SNMP) and Remote Monitoring (RMON). SNMP allows network management programs to trap errors and act on them. RMON-capable components provide the intelligence needed to remotely monitor the devices’ performance and operating environment.

Depending on the intelligence of the components, network management packages can automatically identify the equipment supporting a network. This capability can save time when configuring the network management software. It also simplifies the task of keeping the network management software current as the network expands or the configuration changes. With intelligent devices on the network, sophisticated network management software can monitor key internal functions of the devices, including:

  • Fan speed — monitoring the speed that components run at can provide early warning to impending failure if the component speeds up or slows down.
  • Internal temperature — many failures result from overheating, which damages components. By tracking internal temperatures, network management software can alert the administrator to an impending problem.
  • Bad packet rate — by monitoring the frequency of bad packets sent by a network interface card and comparing this count to a historical profile of that card’s performance, network administrators can replace weak cards before they fail.
  • Latencies — if the latencies start to increase in hubs and switches without an increase in traffic, it may signal a problem with the line or the component. This allows the administrator to preempt problems.
  • CPU utilization — tracking server usage allows network administrators to better balance loads and schedule upgrades before the growth of traffic overwhelms the server.
  • Hard disk and controller performance / errors — nothing is more disheartening than a total disk crash. Some network management software tracks performance and errors on the disk subsystem, which allows the network administrator to swap out drives before they fail.
  • Memory errors — collecting data on memory errors on workstations and servers allows network administrators to repair faulty systems before they create a major problem.

Latest Articles

Follow Us On Social Media

Explore More