BGP (Border Gateway Protocol) is the de facto routing protocol used to route data across the internet and between autonomous systems (ASes). It is a path-vector protocol, which means it uses a list of AS numbers to maintain and share routing information across large networks. BGP is responsible for making inter-domain or inter-AS routing decisions and is designed to handle complex routing policies and large-scale networks.
Key Features of BGP
- Path-Vector Protocol
- BGP uses a path-based approach for routing where routes are represented by a sequence of AS numbers that form a path from the source to the destination.
- Inter-AS Routing
- BGP is primarily used for routing between different autonomous systems (ASes), which are collections of IP networks under the control of a single organization.
- Policy-Based Routing
- BGP allows for policy-based routing decisions. Networks can apply routing policies based on attributes such as the AS Path, local preference, multi-exit discriminator (MED), and more.
- Classless Inter-Domain Routing (CIDR)
- BGP supports CIDR, which allows it to handle IP address blocks more efficiently than older protocols like RIP, which used classful addressing.
- Scalable
- BGP is designed to scale well in large networks with hundreds or thousands of routes. It is capable of handling global internet routing.
- Loop Prevention
- BGP uses the AS Path attribute to prevent routing loops. If a router sees its own AS number in the AS Path, it will reject the route.
- Support for IPv4 and IPv6
- BGP operates over both IPv4 (BGPv4) and IPv6 (BGPv6) networks.
How BGP Works
1. BGP Neighbor Relationship (Peerings)
- BGP routers establish neighbor relationships (peerings) with other BGP routers to exchange routing information.
- These peerings can be internal (iBGP) within the same AS or external (eBGP) between different ASes.
iBGP vs eBGP
- iBGP (Internal BGP): Used for communication between routers within the same AS. iBGP ensures that routing information is shared within an AS.
- eBGP (External BGP): Used for communication between different ASes. eBGP routers exchange routes between distinct ASes on the internet.
2. BGP Path Attributes
BGP routes are advertised with a set of attributes that help determine the best route. These include:
| Attribute | Description |
|---|---|
| AS Path | The sequence of AS numbers that the route has passed through. Helps prevent routing loops. |
| Next Hop | The IP address of the next router to forward traffic to. |
| Local Preference | Used to prefer one route over another within an AS. Higher values are preferred. |
| Multi-Exit Discriminator (MED) | Suggests the preferred path when multiple entry points exist between two ASes. |
| Origin | Indicates how the route was learned (from IGP, EGP, or incomplete). |
| Community | A tag used to apply routing policies across BGP routers, such as identifying specific groups of routes. |
3. BGP Update Messages
BGP routers exchange Update messages to advertise routing information. These updates include:
- Network Prefix: The destination network.
- AS Path: The list of AS numbers traversed by the route.
- Next Hop: The next router to reach the destination.
- Path Attributes: Additional attributes to influence route selection (Local Preference, MED, etc.).
4. BGP Route Selection Process
When a BGP router receives multiple paths to the same destination, it uses the following criteria to select the best route:
- Highest Local Preference: Routes with the highest local preference are preferred.
- Shortest AS Path: The path with the fewest AS hops is preferred.
- Lowest MED: If the AS Path is the same, the router will prefer the route with the lowest multi-exit discriminator (MED).
- EBGP over IBGP: Routes learned from external peers (eBGP) are preferred over those learned from internal peers (iBGP).
- Older Routes: If all else is equal, the router will choose the oldest route (the one that was learned first).
BGP Route Types
BGP classifies routes into several types:
- Network Routes
- Routes learned from BGP peers to reach a particular destination network.
- Aggregate Routes
- A summary route that represents multiple specific routes. Used to reduce the size of the BGP routing table.
- External Routes
- Routes learned from other ASes (via eBGP).
- Internal Routes
- Routes learned within the same AS (via iBGP).
- Locally Originated Routes
- Routes created locally by the BGP router itself, usually to advertise networks directly connected to it.
BGP States
BGP establishes a connection using the following states:
- Idle
- The initial state, where the BGP process has not yet established a connection.
- Connect
- The router attempts to establish a TCP connection to its BGP peer.
- Active
- The router is actively trying to establish the connection but has failed to connect yet.
- OpenSent
- The router has sent an OPEN message to its peer and is awaiting acknowledgment.
- OpenConfirm
- The router is waiting for a KEEPALIVE message after the OPEN message has been acknowledged.
- Established
- The connection has been fully established, and BGP routes are exchanged.
BGP Routing Table
A BGP routing table stores all known routes and their associated attributes. The BGP table is distinct from the IP routing table, as it contains both valid and inactive routes, as well as best routes and backup routes.
BGP Routing Table vs IP Routing Table
- BGP Routing Table: Contains all routes advertised by BGP peers, including non-selected paths.
- IP Routing Table: Contains only the best path for each destination, which is selected by BGP.
BGP vs OSPF vs RIP
| Feature | BGP | OSPF | RIP |
|---|---|---|---|
| Protocol Type | Path-Vector | Link-State | Distance Vector |
| Metric | AS Path, Local Preference, MED, etc. | Cost (based on bandwidth) | Hop Count |
| Network Size | Internet-wide (Inter-domain routing) | Large enterprise networks | Small to medium-sized networks |
| Convergence Speed | Slow (compared to OSPF, EIGRP) | Fast | Slow |
| Scalability | Very High | High | Low |
| Loop Prevention | AS Path (explicit) | SPF Algorithm | Split Horizon, Poison Reverse |
BGP Advantages
✔ Scalable – Can handle large networks and is suitable for global routing on the internet.
✔ Flexible – Supports policy-based routing based on various attributes like AS Path, local preference, MED, and more.
✔ Supports CIDR – Efficient in handling large blocks of IP addresses using CIDR.
✔ Loop-Free – Ensures no routing loops through the AS Path and loop detection.
BGP Disadvantages
❌ Slow Convergence – BGP can take time to converge, particularly during network changes.
❌ Complex Configuration – Requires careful configuration, especially with policy-based routing and multi-hop BGP.
❌ Resource Intensive – BGP routers need significant CPU, memory, and bandwidth to store and process routing tables.
❌ Susceptible to Misconfigurations – Incorrectly configured BGP policies or filters can cause routing issues and traffic misdirection.
Use Cases for BGP
- Internet Backbone Routing
- BGP is used by ISPs and large organizations to manage internet traffic between different ASes.
- Multi-Homing
- BGP is used for multi-homing to provide redundant internet connections from different ISPs, enhancing reliability.
- Traffic Engineering
- Through BGP attributes like local preference, AS Path, and MED, BGP allows for fine-grained traffic management.
- Peering Agreements
- ISPs use BGP for peering agreements, allowing them to exchange traffic with other networks in a mutually beneficial way.
Conclusion
BGP is a powerful, scalable routing protocol that enables inter-domain routing on the internet. It is essential for global routing and managing complex routing policies. Despite its complexity and slower convergence speed, BGP remains the backbone of the internet’s routing infrastructure.
