VNIP collects WAN topology information and simplifies the configuration of frame relay management systems. The protocol lets a device identify the remote equipment found on the opposite end of a PVC. Because VNIP automatically discovers remote devices and maps DLCI and IP addresses on to PVCs, it helps you avoid manual configuration errors. VNIP also provides delay measurements between WAN interfaces. In contrast, common LAN-based IP ping tests may deliver inconsistent results.

The T-FRAP can be configured via a keypad interface, a directly connected terminal or telnet. Performance statistics also appear at this interface, though it's more likely you'll use SNMP to collect this information. SNMP management is resident in all T-FRAP models, accessible through the DLCI techniques described or the unit's SLIP (Serial Line Internet Protocol) interface. The T-FRAP doesn't support direct Ethernet connections, unlike some other manageable CSU/ DSUs, but neither does it require these interfaces for manageability.

For monitoring frame relay WAN performance, Sync offers the Envisage network management system, a Windows-based SNMP console that also performs CSU/DSU configuration tasks. While other frame relay management packages cost $10,000 or more just for starters, Envisage is less than half that price. The software excels at configuring Sync's products and monitoring the network's health, but it lacks extensive reporting capabilities. According to Sync, these trend and reporting deficiencies will be addressed in add-on products that are expected to ship later this fall; they also can be addressed with available third-party SNMP reporting packages.

Under the Covers Envisage consists of two components. The first module, ControlSight, manages device configurations and displays real-time alarms. The second module, InSight, is a Java-based application that provides extensive detail on current frame relay service levels and profiles the traffic being delivered over the WAN. In my tests across a four-site WAN, InSight correctly reported traffic characteristics, utilization, network delay and availability.

When purchased with the Pro-Act software option, the T-FRAP collects details on IP, IPX, SNA, ARP (Address Resolution Protocol) and user-definable protocols. Short-term statistics collect all protocol data on as many as 50 PVCs. Long-term statistics are held for eight protocols, with up to 96 intervals being maintained on the unit--a reasonable but not overly generous amount of data. How often this information must be collected by Envisage depends on how frequently the T-FRAP collects data from the network. If the T-FRAP obtains its data once every minute, the network management system should collect data roughly every hour and a half. If you can settle for less granularity (3,660 seconds is the longest available setting), data could be collected as infrequently as every four days.

Getting Better All the Time Using ControlSight, I updated a pair of the T-FRAPs to a beta release of version 3 code--a trivial and intuitive process. This version of the NMS software (1.101 Build 4) wasn't able to recognize version 3 features such as PVC Loopbacks, but it still was able to manage the remote devices. ControlSight can also launch a telnet session, letting an operator access features found only via the command-line interface on the CSU/DSUs.

The T-FRAP 3.0 introduces some features that you won't find in frame relay-aware CSU/DSUs or management products. Individual PVCs can be looped back for diagnostic purposes, adding to Sync's already extensive loopback capabilities. Even more significant, the T-FRAP can stress-test an individual PVC, running traffic up to 110 percent of CIR (Committed Information Rate) and reporting loss and delay. Thus an installer can verify CIR levels right at the CSU/DSU, before a router or FRAD is installed.

Also new to version 3 is Quick Install, which dramatically reduces the effort to configure individual units. During testing, I answered prompts for an IP address and basic T1 parameters, and the T-FRAP autoconfigured itself to the network, automatically detecting the LMI type and discovering peer devices on the network. Next, I fully configured the T-FRAP through ControlSight by using an existing device configuration as a template and downloading it over the network. This technique is practically effortless, and it's easy to keep configurations consistent.

After the T-FRAP automatically discovered active PVCs and the network began to pass traffic, I ran some PVC loopbacks from the command line. This nondisruptive loopback indicated that the PVC was correctly passing traffic end-to-end, reported the amount of traffic delivered (both marked for discard and not) and calculated the delay across the WAN.

At a lower level, T-FRAP version 3 allows installers to monitor signals on the DTE (data terminal equipment) port and troubleshoot problems with cabling to local equipment. EIA signals such as DTR (Data Terminal Ready), RTS (Request to Send), CTS (Clear to Send), DCD (Data Carrier Detect) and DSR (Data Set Ready) can be monitored at the command line. As another new feature, DCD is supplied to the router during LMI Sourcing, eliminating the need to disable DCD monitoring on the router for it to work.

Send your comments on this article to David Willis at dwillis @nwc.com