To view the Report card.
These baseline reporting systems monitor your network, analyze traffic and report trends and exceptions. Unlike traditional Simple Network Management Protocol (SNMP)-based network management methods, these pr ovide flexible polling, event-correlation engines, active topology maps to represent polled status and very little canned intelligence. They rely on SNMP polling, but add reports on specific device performance in relation to average and maximum performance limits. These reports not only trace what happened the day and week before, but also attempt to peer into the future to predict what lies ahead.

To be successful, these products must be flexible and intelligent enough to adapt easily to different networks. During our performance testing, Network Health's predefined reports excelled in this regard. The product let us make the changes our network needed with a minimum of hassle, and thus receives our Editor's Choice award. Kaspia's Monitoring System version 1.0 release was easier to get up and running--partly because it runs on Microsoft Windows NT server--but despite strong Web-based reports, it wasn't as flexible. And NetworkPM, more of a diagnostic tool, offers strong ad hoc SNMP reporting but scant bas elining, compared to the other two products we evaluated.

Strangely enough, all three lack graphical topology map representations that indicate which segments and nodes have problems (Kaspia's product offers a topology map, but not for troubleshooting purposes). Traditionally, network management systems autodiscover the network, displaying blinking icons and colored lines to represent the logical connectivity via a multilayered topology map. Although the products we tested have an online polling engine, they all rely solely on their reports--rather than on a graphical topology feedback--to convey network status, missing a powerful means to communicate how the network is doing and saddling the reader of the report with the burden of knowing network topology.

Concord Communications Network Health

Wake up and smell the network. When it comes to brewing network knowledge, Network Health's careful blend of high-level and detailed reports poured out the richest network information. Flexible scheduling, predefined modifiable reports and the widest choice of output formats make this product worth its higher price.

The LAN/WAN reports bundle the Overview, Exception, Supplemental, Bandwidth Trend and Multi-Variable Trend network nodes. We scheduled, customized, printed and published to the add-on CERN Web server, Web Link, over a period of weeks to let the baseline averages mature.

The Overview report information was both obvious and mysterious initially. We understood the Daily, Weekly and Hourly Network Volume bar charts, which displayed aggregate network bandwidth as a percentage or number of bytes on a daily and weekly basis. Even the bandwidth trend line that overlaid the Hourly report seemed fairly straightforward, with its comparison of current and historical traffic volume. Another obvious report, Volume Leaders--which displayed, in descending order, the top 10 nodes according to aggregate network bandwidth, again as a percentage or number of bytes--helped us avo id consulting the manual and seemed ideal for management.