Visual UpTime™ and Trouble Sectionalization

Daily ops staff at the NOC watch the Visual UpTime™ Event Processing window to monitor access line status. The ops staff indicates that the added dimension UpTime™ brings to the daily operational practice of alarm monitoring is the timely and accurate isolation and identification of the network component that is in trouble or has failed.

Using only the ping connectivity application, the ops staff can identify if a router is operational and reachable using IP level protocols. But when a router ceases to respond to pings, the ops staff must manually probe adjacent network elements (routers, certain CSU’s) in the network using SNMP NMS "MIB walks" and standard network utilities such as ping and traceroute. If a brokerage router cannot be reached using ping over the FR network, the ops staff had no straightforward way to determine whether the router, CSU, or FR access line is out of service or mis-configured. All these activities were labor-intensive, costly, and prolonged service outage for a mission-critical application.

By adding Visual UpTime™ to existing daily operations, Frame Relay is no longer an opaque "cloud". All the individual components that comprise each Frame Relay connection - router, serial interface, DSU/CSU, telco DDS or T1 facility - are visible and can be managed individually. Armed with this added ability to view components rather than a cloud, operations staff now first consults the Visual UpTime™ Event Processor to determine the root cause for access line trouble or failure. The example in Figure 5, extracted from the subject company’s network performance database, illustrates how Visual UpTime™ reports such events to the customer.

Figure 5. Access line failure event notifications

Once an event is noted, the NOC staff consults the Troubleshooting window for further analysis. Here, staff can view signal and error levels to ascertain whether the access line is OK, degraded, or failed.

The subject company’s investment in configuring Visual UpTime™ with complete naming, including telco circuit ID’s, now pays off, since troubled access lines can be accurately identified to the telco. Location information is also readily available to identify customer(s) impacted by each event.

The ability to isolate and quickly associate a network fault with readily identifiable externally-visible WAN services and facilities is what distinguishes service level management from network management.

The majority of the subject company’s ASE’s are configured to generate events upon detection of access line signal loss. In our opinion, the company will gain an even better understanding of traffic behavior over their WAN circuits when the daily ops staff expands their use of statistical threshold events to include:

• Channel throughput and utilization
• PVC throughput and utilization

This is illustrated in Figure 6, where two of the ASE’s configured to monitor PVC utilization thresholds report that traffic across these DLC’s exceeds the specified threshold.

Figure 6. Traffic Threshold Event Notifications

Monitoring these events will prove especially useful on T1 Frame Relay circuits where the subject company serves 40 or more brokerage locations. Such utilization threshold notifications will provide advance warning when these high-priority circuits are in danger of becoming oversubscribed. The subject company will also be able to determine whether DDS circuits, as configured, will be sufficient when application traffic delivered to the brokerage locations increases.

ASE’s collect and analyze utilization statistics. Only event notifications are transmitted over the network to the PAM or MIC. SNMP NMS’s don’t poll network elements, and processor cycles of network elements are not constantly diverted from packet-switching to responding to management queries. The ability to monitor performance and notify ops staff of significant events in a non-intrusive , low-overhead manner is a unique and attractive feature of the Visual UpTime™ product.

next...