Packet Design's Route Explorer 3.0 Adds Root-Cause Analysis for BGP Networks; Users Can More Rapidly Diagnose and Troubleshoot BGP Problems

Business Wire, Oct 28, 2004

PALO ALTO, Calif. -- Route Explorer 3.0, the newest version of Packet Design's route analytics system, incorporates a BGP root-cause analysis capability that lets service providers and enterprises quickly determine what triggered millions of potentially troublesome events in their BGP (Border Gateway Protocol) networks.

The new BGP root-cause analysis technology allows Route Explorer to provide an accurate view of the service provider's or enterprise's inter-domain layer 3 topology at any point in time, analyze a virtually unlimited number of BGP messages over any selected time period, and determine the causal network event(s) (e.g., a flood of a million BGP messages might be analyzed as a single peering failure). With this knowledge, organizations can more rapidly and effectively diagnose BGP problems and see how routing changes from other providers are impacting their own network.

Route Explorer 3.0 and a new BGP analysis module that includes the root-cause analysis technology will be demonstrated at the ISPCON show Nov. 3-5 at the Santa Clara (Calif.) Convention Center, booth #503. Also on display at ISPCON will be the recently introduced Route Explorer/MPLS VPN, which lets service providers monitor individual customer BGP/MPLS VPNs for proper route distribution - a major step in ensuring virtual private network reachability and privacy. Both products began shipping in late October.

BGP Root-Cause Analysis: Speeding Detection, Diagnosis of BGP Problems

"Networks based on BGP have traditionally been difficult to diagnose and troubleshoot because this 'chatty' protocol can generate thousands, even millions, of updates following a peering loss or other significant routing event," said Jeff Raice, Packet Design executive vice president of marketing and business development. "With today's large and complex networks, making sense of these events is beyond human capability. As a result, network operators are left reacting to customer complaints, when they would prefer to address issues proactively, before complaints surface."

Packet Design's new BGP Root-Cause Analysis software has two components: a BGP RIB (routing information base) visualization feature, and a dynamic root-cause analysis capability that can analyze millions of BGP events in seconds and animate those events on a topology map.

The BGP RIB visualization tool processes the data gathered by Route Explorer's BGP route analytics engine and creates a visual representation of the service provider's inter-domain BGP network, showing its peer autonomous systems (ASs) and the significant ASs connected to its peers and beyond. By showing where all BGP prefix announcements are coming from, this AS topology map allows service providers to view at a glance their routing policies (i.e., how they are connected to other providers), validate existing peering arrangements and plan future ones. It also lets them rapidly diagnose problems such as misconfigured community tags and unexpected or unwanted backdoor paths.

The root-cause analysis feature quickly analyzes huge numbers of BGP routing messages to single out the key routing events that were the source of major routing activity. Using Route Explorer's History Navigator, the user can select a time period characterized by spikes in routing events or an unusual BGP routing activity level. A statistical algorithm developed by Packet Design extracts the large-scale structure of BGP event streams, examining millions of BGP events in seconds and identifying the single or multiple events that triggered them, even if those events are several "hops" away from the user's network. Root-cause events such as peering flaps or MED (Multi Exit Discriminator) oscillations take days to find (if they are ever found); even then, the user cannot determine why they occurred. With Route Explorer, the user can select one such event and immediately see a dynamic BGP topology map showing routes being withdrawn from one AS or peer and moving to another. For example, a service provider could easily see that a peering flap between providers several hops away from its network was causing a direct peer to periodically withdraw and announce the routes. This powerful visualization technique provides quick answers to previously unanswerable questions, such as, "What happened?", "Where did it happen," and "How does it affect me?" The dynamic visualizations can be saved and emailed between service providers or by enterprises to their providers, rapidly conveying the nature of the problem and coordinating fast inter-organization problem resolution.

Route Explorer MPLS VPN: Ensuring Reachability, Privacy, Site-to-Site Policy

MPLS VPNs based on the IETF RFC 2547bis standard have been embraced by service providers as a way to give enterprise customers the flexibility of provider-managed layer 3 connectivity among their distributed sites. But, without layer 3 visibility, the providers have lacked knowledge of when service-affecting routing errors occurred and how to fix them.