I tested several new capabilities of a beta version of WiseWan version 2 for Solaris, using several WiseWan 200 probes at the MCI Developer's Lab in Richardson, Texas, and I liked what I saw. Version 2 introduces many significant enhancements to the product, including per-stream bandwidth limiting, response-time monitoring and a port to Windows NT. Some of these items were still in development when I saw the product. If NetReality can deliver on its plans, WiseWan 2 will set a new capability standard for WAN management systems.

Two Tastes Go Great Together QoS (quality of service) management and frame relay monitoring are a natural fit. Both are driven by the high cost of WAN bandwidth and the potential for its misuse. Both often require devices at the same location, managed by similar consoles. Why not combine them into a single management station using the same instrumentation? That's the genius behind WiseWan.

WiseWan's frame relay monitoring features meet or exceed those of several other management systems. The WanXplorer application provides real-time event monitoring with intelligent alarm correlation, automatically updating old conditions that have since been corrected. And it provides reports based on real-time and historical data, concentrating on typical frame relay metrics, including line health and utilization and PVC-level traffic details.

WanXplorer features a wide range of reporting capabilities. Real-time reports show conditions as they occur based on per-minute polling of remote probes. History reports display longer intervals, while Typical reports use daily or weekly averages for simple trend analysis. Much of the data is displayed graphically in a variety of user-selectable chart styles. WanXplorer does not offer the detailed trending capabilities found in systems such as Concord Communications' Network Health, but can provide data to these systems if you need it.

Standard reports show line health and usage, top DLCIs (Data Link Connection Identifiers) based on a variety of indicators, and DLCI health and utilization. Network and application protocol distribution displays the relative use of up to nine protocols. Top hosts and conversation display the network's biggest bandwidth hogs.

Some events and reports were not available in the beta version I tested, but are promised by the time version 2 ships. Most noticeably absent were alarms and reports based on frames marked as Discard Eligible (DE). DE marking is a critical metric for verifying a PVC's CIR (Permanent Virtual Circuit's Committed Information Rate), and is a cornerstone of most WAN carrier SLAs (service-level agreements). Surprisingly, alarms based on congestion notifications (FECN and BECN) were operable even though these conditions don't reveal actionable information to the WAN manager.

Current omissions aside, the WanXplorer client application is useful right out of the box. By default, reasonable alarm thresholds are set for most critical events, such as line up/down, high/normal line utilization and PVC state changes. Most metrics have a "rising" setting, which creates a red alarm, and a corresponding "falling" setting, which resolves the original alarm to gray. These thresholds may be easily tuned to meet your particular needs.