The worldwide public internet has long depended on the Border Gateway Protocol (BGP) for exchanging routing a reachability information between administrative domains (autonomous systems / ASs). With the advent of MPLS VPNs, we have not only seen BGP become prevalent in the Service Provider access layer but increasingly become the core routing protocol for private enterprise wide area networks as well. More recently, we have witnessed the broad acceptance of BGP as a data center fabric control plane protocol, which will ultimately result in the most dramatic increase of BGP endpoints so far.

Given how important this protocol has become to the overall operation of network infrastructure and services, operations teams must have visibility into all aspects of active BGP sessions, enabling them to spot issues or abnormalities before they escalate into complete outages.  Until now that has only been possible through BGP route servers or techniques such as SNMP polling or CLI scraping, which provide extremely limited insight and are inherently inefficient and therefore ineffective. While the most granular information can be obtained through CLI scraping, this technique relies on periodically polling each device in the network. Not only does this have a detrimental effect on platform performance but it results in the distinct possibility of critical changes being missed in-between polls. With the need to have a daemon running on each BGP router in your network, CLI scraping also exposes serious security vulnerabilities.

The introduction of the BGP Monitoring Protocol (BMP) totally changes the analytics landscape, rendering those inefficient polling-based techniques effectively obsolete. By presenting the exact information the BGP algorithm used to make its best path decisions, BMP provides external applications the most comprehensive insight possible into the real-time (and historical) operation of a network. BMP implementations comprise an agent client, which resides in the routing component and a collector server. When BMP is running on your BGP router, it generates events, including updates on BGP session state and prefix activity, that can be collected centrally and periodically sends out stats reports such as the number of prefixes rejected by the inbound policy.

In this short video, we take a look at the Metaswitch implementation of BMP -- available as part of the Metaswitch Network Operating System (NOS) Toolkit -- and how it can be used in a network to provide a real-time graphical view of the operation of a BGP network infrastructure. 

To learn more about the Metaswitch NOS Toolkit, download our NOS Cookbook