Introduction
Scope of Evaluation
What is Visual UpTime™?
Case Study Overview
Daily Ops & UpTime™
Troubleshooting FR
Managing FR Traffic
Tip of Iceberg
Conclusions
Contact Visual Networks

Visual UpTime™:
A Core Competence Product Evaluation
The Visual UpTime™ WAN Service Level Management System

Enterprise network management of WAN services such as Frame Relay is typically performed from a central, commercial SNMP network management system (NMS). By polling software agents in network elements, including routers, FRADs, and certain WAN DSU/CSU equipment, enterprise network administrators hope to obtain status and performance information necessary to determine whether their WAN services are operational and performing as anticipated.

There are several problems with this approach:

  1. Scalability. The use of polling to monitor the operational status of large numbers of WAN circuits has proven to be difficult. Administrators have turned to alarm processing, but generic SNMP traps offer too little information to be useful for monitoring service levels. Conventional SNMP agents cannot monitor performance and transient error conditions before they issue alarm messages.
  2. Integration. Commercial NMSs require customization and frequent polling to collect information from the various network elements needed to analyze WAN performance. Traffic associated with collection of performance data can impose a constant load on a network. This is undesirable for WAN services where subscribers pay based on use or information rate commitments.
  3. Visibility. Commercial NMSs cannot collect sufficient data from conventional SNMP agents to perform centralized real-time analysis of WAN circuit traffic. Network administrators must capture and analyze traffic on LANs adjacent to WAN circuits to identify higher level protocol problems. "Sniffing" at a LAN interface does not always provide enough information to solve a WAN problem, and requires that expert staff be present at the site experiencing the problem.

Network administrators need proactive and real-time tools to deal with these problems. They need tools that provide immediate notification of service failures, tools that accurately identify the causes of service failures. They need tools that constantly monitor service levels, maintain detailed performance data for all WAN circuits, and provide readily accessible analysis and reports to corroborate or refute claims that service level commitments are being met. This is the essence of WAN Service Level Management.

Visual UpTime™ WAN Service Level Management System

The Visual UpTime™ WAN Service Level Management system complements existing network elements, SNMP agents, and network management systems to support WAN service level management.

  • At each access line in a WAN, an intelligent agent called an Analysis Service Element (ASE) gathers WAN service status and statistical information.
  • Data are gathered from all the ASE's via standard SNMP and bulk file transfer protocols to form a detailed and long-term network performance database.
  • Toolsets accessed through a console application provide the ability to examine and manipulate the network performance database.

Visual UpTime™ users can choose an integrated version of the console application called a Management Integrated Console (MIC), or a distributed, client-server management product, in which a Performance Archive Manager (PAM) provides client access to the management toolset from Platform Applicable Clients (PACs). Both the PAM and MIC are pre-installed and come ready to use on a Pentium PC operating under Windows NT Server 4.0. PACs are available for many UNIX OSs and any Microsoft Windows 32-bit OS.

Analysis service elements are intelligent agents that continuously monitor activity on a WAN access circuit. Unlike conventional SNMP agents, ASE's do not generate a constant stream of management traffic from network management agent to NMS. ASE's are proactive monitors: they generate alarms when user-defined thresholds for performance-related events are crossed. This reduces or in eliminates management overhead associated with the "polling" of network elements by a conventional SNMP NMS.

The heart of the Visual WAN SLM system is the MIC or PAM. The MIC (PAM) collects and processes network operational data, and prepares the data for reporting and display. The console application manages the network performance database and supports toolsets for monitoring, troubleshooting and analyzing WAN services.

The data gathered from ASE's forms a highly detailed and long-term network performance database containing physical access circuit, link layer and payload analysis data. The data collected by ASE's are stored in a two-day database "buffer". Aggregated data is summarized and retained in this database, limited only by allocated database storage. This Microsoft SQL Server 6.5 database is accessed through a toolset provided by a console application.

The toolsets available from the MIC or PAC console application include:

  • Network Configuration. A toolset used to identify WAN networks and network elements that comprise the WAN network to be managed.
  • Performance Monitoring. A toolset used to identify WAN performance events to be monitored and to configure ASE threshold values for these events.
  • Event Processing. A toolset used to provide real-time event and alarm display and to manipulate logged ASE-generated events and alarms.
  • Troubleshooting and Statistical Analysis. View real-time and historical statistical information by access line, access channel, and PVC/Protocol.
  • Network Traffic Capture and Decoding. The protocol analysis or "packet-sniffing" functionality of the ASE.
  • Planning and Reporting. Create reports from the long-term database maintained by the PAC/PAM or MIC.

next...