BGP and routing policies
Layer 3 networking with Border Gateway Protocol (BGP) offers superior scalability, resiliency, and control compared to point-to-point Ethernet solutions. By leveraging dynamic path selection and policy-based routing, BGP enables intelligent traffic engineering, automatic failover, and multi-site connectivity across complex or multi-provider environments. This makes Multi-Cloud Gateway ideal for enterprises that require redundancy, optimized performance, and flexible routing strategies beyond simple point-to-point transport.
BGP sessions
BGP is dynamic routing that propagates route table updates, exchanging routing information between peers to provide flexibility.
Multi-Cloud Gateway BGP session features:
- Directly connected neighbors supported. Lower hold time values negotiated during session establishment.
- Default Lumen ASN is AS1.
- Bidirectional Forwarding Detection (BFD), a network fault detect protocol that detects path failures between BGP neighbors, can be enabled to quickly detect failures and facilitates faster re-convergence time. This allows a BGP peer relationship to be torn down quickly after notifications from BFD, failing over to another BGP neighbor.
Default interval (time between BFD control packets) is 500 ms and multiplier (number of missed packets before designating session down) is 5. Interval values are 500, 1000, 1500, and 2000. Multiplier range is from 5 to 10.
- MD5 authentication provides additional security.
- Use default originate if you don’t have a default route and Lumen will announce a default route with default originate configuration.
- An autonomous system (AS) refers to either a single network or a group of networks and routers managed by one administrator or administrative team representing a single organization, such as a business division. Each AS receives a globally unique number that serves as its identifier across the internet. ASN support includes configuring local ASNs for individual peers.
Prefix lists
Prefix lists make it easy to configure and manage routing tables. Create a prefix list from IP addresses you use frequently, then reference them as a set in routes instead of referencing them individually.
- Filter lists for route filtering and manipulation.
- Deny all is default.
- Option to advertise all routes to BGP neighbors and advertise all routes to the Multi-Cloud Gateway or the Cloud Interconnect connection to your port.
- Indicate the type of match (exact or exact plus longer) to perform on a prefix.
- Import and export policies to control routes to and from.
- Multi-Exit Discriminator (MED) influences a BGP neighbor so that it takes a preferred route if the advertising AS is the same for available routes and the AS has multiple entry points. A lower MED is preferred. Use MED to avoid equal-cost multipath (ECMP) behavior. This attribute breaks ties between two routes when local preference, AS path, and origin are the same.
- Add AS prepend to set route advertisement priorities. It repeats the originating AS number multiple times to artificially lengthen a route’s AS path. This makes a route less preferred since BGP prefers routes with shorter AS paths.
Static routing
Static routes are manually configured network paths that define how traffic should be directed from one device to another. Static routes provide a fixed, predictable route for data to follow. This makes them particularly useful in smaller or stable networks where simplicity and control are priorities. Unlike dynamic routing, static routes require manual updates if the network topology changes.