NetMRI's forté lies in problem analysis. Each NetMRI appliance, which typically connects to a central pathway on the network and can monitor as many as 2,000 devices, contains a large body of expert system knowledge developed by Netcordia. The company says the appliance can tap into these expert rules to assess, audit and actively detect more than 200 system-configuration and network QoS issues.

NetMRI promptly and unerringly pinpointed and alerted us to all the network, application and server problems we threw at it -- outages, slowdowns and overburdened devices, degraded links and overtaxed computers. NetMRI notifies administrators -- and others via escalation, if the problem persists -- using e-mail and pager. For problems susceptible to automatic repair, NetMRI can run scripts and external programs, reboot a server and restart a failed background service.

While other appliances typically require ASCII-terminal-based initial configuration, NetMRI uses a browser-based wizard for painless and quick configuration.

NetMRI's discovery feature quickly and efficiently gave us an accurate inventory of our network devices and computers. Starting with IP address ranges we specified, a seed router we pointed at or Comma Separated Value data we imported, the NetMRI unit identified our routers, switches, servers and clients. At a frequency you specify, NetMRI automatically looks for and catalogs new devices on the network.

NetMRI's monitoring function has a sophisticated threshold facility that an IT manager can use to precisely express unacceptable activity levels on the network at specific times of the day for specific days of the week. NetMRI's Routing Neighbor Analysis feature is especially helpful in that it alerts you to changes in live routing pathways. This feature not only informs you about WAN link outages, but also specifies the links that will shoulder the new traffic burdens as a result of those outages.