In addition, SLM better utilizes enterprise MIBs on routers and switches from Cisco Systems and assesses network health for Cisco devices. However, I wish SLM would at least report rudimentary health and availability information on my Sun Microsystems SPARC Solaris servers. Network Associates says it plans to add support for more devices in a future release.

I added support for end-user monitoring by installing VitalSigns' Net.Medic on a few desktops and Vital.Analysis on the server. Net.Medic collects information about client machine and Internet server performance, as well as all of the in-between plumbing. I configured the clients to upload the information to my Vital.Analysis server once a day. The server then transports that information to the SLM database, which, in turn, churns out network health and availability from the user's perspective.

A few days later, I fired up a Web browser and connected to the main SSDS server, which presented a menu of the various reports, such as SLM and the PM products' exception reports, and a link to the SniffMaster browser. The SLM module offers general availability and health reporting on routers, switches, networks, clients and servers. The PM products are configured to collect detailed information on specific network devices, including routers, switches and frame relay devices. In the lab, SLM simply told us that a router was not 100 percent healthy, while RouterPM told us that the router was running short on memory buffers.

Currently, SLM supports only Windows NT and Novell NetWare servers. When I selected the general report, I saw the overview information in two categories: NetWare and Windows NT. From there, I clicked on the NetWare hyperlink to look at individual server statistics. Clicking on our production NetWare server, I obtained server health and availability graphs. Your browser must support PNG for these graphs. However, Network Associates says it's planning to switch back to JPEG format because of poor browser support for PNG.

Novell's NMA provided the additional information I needed about our NetWare servers. After completing a data-collection cycle, I looked at trends in our servers, such as licensed connections, memory use and server utilization. In the future, this information will help me determine when I will need to think about purchasing more memory or connection licenses.

Desktop client reports also are integrated into SLM using Net.Medic on the client and Vital.Analysis on the server. As in a standard Vital. Analysis installation, Net.Medic collects statistics at the desktop. The client collects extensive information about its Web browsing, including throughput and latency on the Internet, intranet and local LAN. This information is uploaded at regular intervals (for us, once a day) to the Vital.Analysis server, where it is analyzed. Unique to the SSDS integration, Vital.Analysis transmits this information to the SLM server.

Installing the Net.Medic clients was an easy task, but getting them to divulge their information to the Vital.Analysis server was more involved. I would like to see Network Associates add more diagnostic tools that describe when and if the data will be uploaded. I blindly made changes to configurations and had to wait 24 hours to see if the changes made any difference.

In addition, if you eventually want data uploaded once a day during off hours, users will need to leave their machines on overnight.

Digging Deeper Beyond the basic, canned SLM reports, the PM products--RouterPM, FramePM, Switch- PM, NetworkPM and HubPM--offer more detailed information, so you can locate potential trouble spots. Although fairly dependent on Cisco infrastructure (support for Bay Network products is in beta tests, according to the company), the PM products monitor such information as router memory use, segment bandwidth utilization, PVC congestion and packet drops.

Within a day of monitoring, we found routers that were low on memory, segments with unusually high collision rate and frame relay PVCs (Permanent Virtual Circuits) that were becoming congested at peak times. In the next version of this software suite, I would like to see consolidated exception reports sent daily via e-mail.

The PM products are completely separate from SLM, so you need to manually enter individual devices into the PM management table. I would like to see better integration into the SLM autodiscovery module. If you find a device in SLM that is also monitored by a PM product, you can click on the device and go to its PM information. However, if I have a device in SLM that isn't monitored in any PM product, I'd like to be able to add it quickly from SLM.