Understanding Border Gateway Protocol (BGP)

The Internet is a decentralized network of interconnected computers and devices that communicate using standardized protocols. It is made up of thousands of independent networks, known as Autonomous Systems (AS), each operated by Internet Service Providers (ISPs), corporations, universities, and other organizations.

The Border Gateway Protocol (BGP) is an essential Internet protocol that ensures traffic gets from point A to point B. With BGP, an AS can determine the most efficient route for a network packet to reach its destination and send it on its way.

How Does BGP Work?

Each AS on the Internet is directly connected to one or more other AS, its “peers” or “neighbors”. These directly linked AS share routing information with one another, including passing on information received from other peers. This peer-to-peer approach allows information on available routes to percolate through the entire Internet via multiple hops.

Every AS on the Internet, through a process called route advertisement,  will share the range(s) of IP addresses that they directly control. For example, “ISP A” will announce that it has a direct route to all of the IP addresses of its customers, including “Bob”, one of its customers.

These route advertisements are shared with ISP’s neighbors or peers, who store them in their routing tables. As a result, these AS will now have a route to these IP addresses. Using our example,  they know that any traffic that they receive for Bob can be sent to ISP A, who will then forward it to the appropriate recipient. These AS will forward new routes to their neighbors until all AS have a route to Bob, and other customers of ISP A.

Since many AS connect to several other AS, there may be multiple routes available from one ISP to another. BGP uses various attributes to select a route, including latency, hop count, and transmission cost. These attributes are included in route announcements, enabling each AS to make its own routing decisions.

Key Features of BGP

BGP is unique among routing protocols because it is designed for inter-domain routing, enabling the exchange of routing information between autonomous systems. Here’s what sets BGP apart from other protocols:

• Path-Vector Protocol: Unlike distance-vector or link-state protocols, BGP maintains a table of AS paths rather than just hop counts or link states, allowing for more flexible and policy-driven routing.
• AS Peering: BGP permits AS to communicate route information to one another. This is accomplished by sharing information with direct peers and propagating these changes across the Internet.
• BGP Attributes: While an AS may prefer the shortest route to a destination, it can also use other factors to help make its selection. Information such as latency or transmission cost is communicated via attributes.
• Route Filtering: BGP allows AS to select the routes that they accept or advertise. Additionally, they can make path selections based on corporate policies and the attributes attached to routes.

How BGP is Used in Internet Routing

BGP is the primary protocol used to route traffic over the network. Any traffic that leaves its source AS will be routed based on information distributed via BGP.

The Advantages and Disadvantages of BGP

BGP has numerous advantages, including:

• Scalability: BGP can handle hundreds of thousands of routes, making it suitable for the vast and decentralized nature of the Internet.
• Efficiency: BGP uses optimized path selection algorithms and minimal bandwidth utilization for route updates.
• Configurability: BGP permits policy-based routing, multi-homed connectivity, and traffic engineering.
• Route Aggregation: BGP allows route aggregation, reducing the number of unique routes to be stored.

However, the protocol also has its downsides, including:

• Memory Requirements: BGP’s support for large routing tables also means that it has higher memory requirements.
• Slow Convergence: BGP can adapt to changes more slowly than other protocols.
• Security Risks: BGP implicitly trusts advertised routes, making it vulnerable to route hijacking and similar threats.
• Suboptimal Routing: BGP may choose suboptimal routes due to prioritizing other factors and its blindness to network congestion.

How BGP Compares to Other Routing Protocols

Routing protocols can be classified as exterior or interior routing protocols. Exterior routing protocols are primarily used for routing information between AS on the Internet. BGP is an example of an exterior routing protocol.

Other routing protocols, such as Open Shortest Path First (OSPF), Intermediate System-to-Intermediate System (IS-IS), or the Enhanced Interior Gateway Routing Protocol (EIGRP) are examples of interior routing protocols. These are designed to route traffic within an AS’s network.

BGP has use cases for internal, as well as external routing scenarios, and can be distinguished by ‘internal BGP’ vs. ‘external BGP’. However, iBGP is rarely used as a primary internal routing algorithm due to its limited scalability and slow convergence. Instead, it’s used as a secondary algorithm supporting OSPF or IS-IS.

Common BGP Configuration Challenges

BGP can be complex to configure manually. Some common issues include:

• Incorrect route summarization causing reachability issues.
• Challenges setting up or maintaining peer relationships.
• Route flapping causing unstable routing.
• Misconfigured filters impacting route propagation.
• Misconfigured BGP attributes impacting route selection.

Effectively Troubleshooting BGP

BGP challenges can be caused by internal and external factors. Some best practices to avoid these challenges include:

• Continuous monitoring to promptly identify and address issues.
• Changing peers as needed due to peering challenges or issues with route data.
• Implementing Resource Public Key Infrastructure (RPKI) to prevent route hijacking.

Integration of BGP in SD-WAN Solutions

Software-defined wide area networking (SD-WAN) separates network control from the underlying transport infrastructure. By leveraging this overlay-underlay separation, SD-WAN enhances performance, security, and flexibility, making WAN management more efficient, resilient, and cost-effective compared to traditional networking.

SD-WAN solutions can implement BGP for both internal and external traffic routing. Major SD-WAN providers, like Cato, have their own infrastructure and operate as an AS. This means that BGP is used to route traffic to the users of their SD-WAN service. SD-WAN providers can also use BGP to implement internal routing. Once traffic enters the provider’s infrastructure, iBGP can be used to direct it to the correct endpoint.

Leveraging BGP for Optimized SD-WAN and SASE Solutions

Cato operates SD-WAN and Secure Access Service Edge (SASE) services on top of a Global Private Backbone. The scale of Cato’s infrastructure makes it equivalent to a Global Service Provider (GSP) and an independent AS within the Internet’s BGP ecosystem.

Cato SASE Cloud uses BGP both internally and externally to ensure secure, efficient routing of traffic to and from its customers. BGP’s dynamic and resilient route selection helps Cato to ensure seamless connectivity and provide reliable and secure data transmission to all of its enterprise users. To learn more about optimizing your network performance and security with Cato, book a demo.