BGP uses several attributes or metrics for the path selection process, this post discusses the BGP attributes used on Cisco IOS-XE routers/switches for path selection and provides configurations examples of commonly used settings.
BGP Attributes
The table below represents the BGP attributes
| BGP Attribute | Description | Category |
| Origin | The source of the path information:
– IGP – Learned by way of network statement. – EGP – Learned from a BGP peer – Incomplete – Learned from redistribution Prefers IGP over EGP routes and EGP routes over routes from incomplete paths.
|
Well-known mandatory |
| AS_PATH | Lists the ASNs in the Path
Each AS prepends its own ASN The AS path describes all the ASA a packet would have to travel to reach the destination IP network. Used to ensure a loop free path. The route with the fewest AS hops is preferred (if used for path selection). If there is a tie with other attributes, MED breaks the tie. |
Well-known mandatory |
| Next hop | Specifies the IP address of the next hop to the destination network. | Well-known mandatory |
| Local preference | The path to use to exit the AS
Not passed on to peers. Default preference is 100 The higher local preference is preferred. Configured on a router, which advertises the local preference to internal iBGP peers. |
Well-known discretionary |
| Multi-exit discriminator (MED) | Informs the external BGP peer the preferred path into the AS when multiple paths exit to the same AS.
Not used with iBGP peers The lowest MED value is preferred, the default value is 0. MED is carried into an ASA but does not leave the AS. |
Optional |
| Community | Groups routes and applies policies or decisions.
NOTE – not used in routing decision process. |
Optional |
| Atomic aggregate | Informs BGP peers the local router used a less specific (aggregated) route to a destination, instead of a more specific route. | Well-known discretionary |
| Weight | Cisco proprietary
Specifies a preferred path is multiple exits exist to a destination. Assigned locally on a router. Weight value ranges from 0 to 65,535 Routes with a higher weight are preferred when multiple routes to a destination exist. Not exchanged in BGP updates. |
Optional |
Cisco routers running BGP use the following algorithms to determine the next path in the following order.
| BGP Best Path Order |
| 1. Highest Weight |
| 2. Highest Local Preference |
| 3. Prefer local originated route |
| 4. Shortest AS_PATH |
| 5. Lowest origin type |
| 6. Lowest MED |
| 7. Prefer eBGP over iBGP |
| 8. Lowest IGP metric to the BGP next hop |
| 9. Oldest path |
| 10. Lowest BGP router ID source |
| 11. Minimum cluster list length |
| 12. Lowest neighbour address |
The following options influence outbound routing decisions
- Weight (Cisco routers only)
- Local preference
- AS Path length
The following options influence inbound routing decisions:-
- AS Path length
- BGP Communities
- MED (Multi-Exit Discriminator)
Load Balancing
By default, BGP does not load balance traffic over multiple links. The maximum-path command must be configured to install more than the first route in the routing table. For multiple routes to be installed the following BGP attributes must be the same:-
- Weight
- Local Preference
- AS Path – both AS number and AS path length
- Origin code
- MED
- IGP metric
The maximum-path command is configured under the address-family.
router bgp 100 address-family ipv4 maximum-paths 2
If the router is connected to two or more routers with a different AS, traffic will not be load balanced, because the AS Path is different. The command bgp bestpath as-path multipath-relax must be used in additional to the maximum-path command.
router bgp 100 bgp log-neighbor-changes bgp bestpath as-path multipath-relax neighbor 1.1.1.2 remote-as 200 neighbor 2.2.2.3 remote-as 300 ! address-family ipv4 neighbor 1.1.1.2 activate neighbor 2.2.2.3 activate maximum-paths 2 exit-address-family
Once configured clear the routes using clear ip bgp * soft and run show ip route bgp to configure two routes to the same destination.
R1# show ip route bgp Gateway of last resort is not set 10.0.0.0/24 is subnetted, 1 subnets B 10.10.0.0 [20/0] via 2.2.2.3, 00:00:12 [20/0] via 1.1.1.2, 00:00:12
Run the command show ip bgp to check the BGP table. From the output below, the route via 1.1.1.2 is the best route, the route via 2.2.2.3 is a multipath route, signified by the *m
R1#show ip bgp BGP table version is 4, local router ID is 192.168.1.1 Status codes: s suppressed, d damped, h history, * valid, > best, i - internal, r RIB-failure, S Stale, m multipath, b backup-path, f RT-Filter, x best-external, a additional-path, c RIB-compressed, t secondary path, Origin codes: i - IGP, e - EGP, ? - incomplete RPKI validation codes: V valid, I invalid, N Not found Network Next Hop Metric LocPrf Weight Path *> 10.10.0.0/24 1.1.1.2 0 200 400 i *m 2.2.2.3 0 300 400 i
Weight
Weight is applied to inbound routes and will influence the outbound path from the local router. The path with the highest weight is preferred. It is NOT passed between BGP neighbours.
ip prefix-list WEIGHTED seq 5 permit 10.10.0.0/24 ! route-map WEIGHTED permit 10 match ip address prefix-list WEIGHTED set weight 123 route-map WEIGHTED permit 20 ! router bgp 100 neighbour 2.2.2.3 route-map WEIGHTED in
Flush the BGP routes and then run show ip bgp. From the output below we can confirm the weight of 123 has been applied to the route via the next hop 2.2.2.3, which is now the best path.
R1#show ip bgp BGP table version is 3, local router ID is 192.168.1.1 Status codes: s suppressed, d damped, h history, * valid, > best, i - internal, r RIB-failure, S Stale, m multipath, b backup-path, f RT-Filter, x best-external, a additional-path, c RIB-compressed, t secondary path, Origin codes: i - IGP, e - EGP, ? - incomplete RPKI validation codes: V valid, I invalid, N Not found Network Next Hop Metric LocPrf Weight Path *> 10.10.0.0/24 2.2.2.3 123 300 400 i * 1.1.1.2 0 200 400 i
Local Preference
The Local Preference attribute controls outbound traffic within the local AS (iBGP). In this scenario two routers within AS 200, the first router uses the local preference to influence the second router in the same AS to route traffic via the first router, rather than direct.
ip prefix-list LOCAL-PREFERENCE seq 10 permit 10.10.0.0/24 ! route-map LOCAL-PREFERENCE permit 10 match ip address prefix-list LOCAL-PREFERENCE set local-preference 123 route-map LOCAL-PREFERENCE permit 20 ! router bgp 200 bgp log-neighbor-changes neighbor 1.1.1.1 remote-as 200 neighbor 4.4.4.4 remote-as 400 ! address-family ipv4 neighbor 1.1.1.1 activate neighbor 1.1.1.1 next-hop-self neighbor 4.4.4.4 route-map LOCAL-PREFERENCE in
Configure next-hop-self is configured on the first router to set the next hop IP address to that of the first router, this will cause reachability issues with iBGP.
From the output of show ip bgp from the second router we can see the local preference of 123 is influencing the next hop, traffic is routed via the first router.
R2(config-router)#do show ip bgp BGP table version is 5, local router ID is 192.168.1.1 Status codes: s suppressed, d damped, h history, * valid, > best, i - internal, r RIB-failure, S Stale, m multipath, b backup-path, f RT-Filter, x best-external, a additional-path, c RIB-compressed, t secondary path, Origin codes: i - IGP, e - EGP, ? - incomplete RPKI validation codes: V valid, I invalid, N Not found Network Next Hop Metric LocPrf Weight Path *>i 10.10.0.0/24 1.1.1.2 0 123 0 400 i * 2.2.2.3 0 300 400 i
Origin
The origin attribute distinguishes between the origin of the how the route was advertised into BGP. There are three methods, IGP (i) learned via a network statement, EGP (e) learned from a BGP peer or Incomplete (?) learned from redistribution from one routing protocol into BGP.
NOTE – EGP is no longer used.
- IGP routes have the highest priority over EGP and Incomplete routes.
- EGP has a higher priority over incomplete routes.
From the output of show ip bgp below, we can determine the origin of the routes for 172.21.0.0/24 is learnt from 2 peers. The best route is via 4.4.4.2 with the origin being IGP (i), due to being configured using the network command. The second route is via 5.5.5.3 and an incomplete route, due to being redistributed into BGP.
R4#show ip bgp BGP table version is 14, local router ID is 10.10.0.1 Status codes: s suppressed, d damped, h history, * valid, > best, i - internal, r RIB-failure, S Stale, m multipath, b backup-path, f RT-Filter, x best-external, a additional-path, c RIB-compressed, t secondary path, Origin codes: i - IGP, e - EGP, ? - incomplete RPKI validation codes: V valid, I invalid, N Not found Network Next Hop Metric LocPrf Weight Path *> 172.21.0.0/24 4.4.4.2 0 0 200 i * 5.5.5.3 0 0 300 ?
AS Path
AS Path is used to influence inbound routing, the local router prepends the AS Path to advertised outbound routes to a BGP peer to make the path less desirable.
route-map AS-PATH permit 10 set as-path prepend 1234 1234 1234 ! router bgp 100 neighbor 1.1.1.2 route-map AS-PATH out
Run show ip bgp on the remote router, we can see the received route 192.168.250.0/30, the AS Path has been prepended with 1234 x 3 and thus less desirable, the route with a short path via 5.5.5.3 is installed as the best route.
R4#show ip bgp BGP table version is 11, local router ID is 10.10.0.1 Status codes: s suppressed, d damped, h history, * valid, > best, i - internal, r RIB-failure, S Stale, m multipath, b backup-path, f RT-Filter, x best-external, a additional-path, c RIB-compressed, t secondary path, Origin codes: i - IGP, e - EGP, ? - incomplete RPKI validation codes: V valid, I invalid, N Not found Network Next Hop Metric LocPrf Weight Path *> 10.10.0.0/24 0.0.0.0 0 32768 i * 192.168.250.0/30 4.4.4.2 0 200 100 1234 1234 1234 i *> 5.5.5.3 0 300 100 i
MultiExit Discriminator (MED)
When multiple paths exist into an AS, MED can be configured outbound on a router to influence the path inbound to the AS. The MED metric is not propagated to any other AS’s (other than the neighbour AS). The lowest MED metric is the preferred path.
A prefix-list identifies the local networks to match, the route-map matches the prefix-list and sets a metric to be applied to that route(s). The route-map is applied to a neighbour.
ip prefix-list MED seq 5 permit 10.10.0.0/24 ! route-map MED permit 10 match ip address prefix-list MED set metric 12345 router bgp 100 neighbor 1.1.1.1 route-map MED out
Once configured the output of the upstream router confirms only the route to 10.10.0.0/24 has a metric of 12345 and is NOT the best path.
R1#show ip bgp BGP table version is 41, local router ID is 192.168.1.1 Status codes: s suppressed, d damped, h history, * valid, > best, i - internal, r RIB-failure, S Stale, m multipath, b backup-path, f RT-Filter, x best-external, a additional-path, c RIB-compressed, t secondary path, Origin codes: i - IGP, e - EGP, ? - incomplete RPKI validation codes: V valid, I invalid, N Not found Network Next Hop Metric LocPrfWeight Path *m 4.4.4.0/24 2.2.2.3 0 200 400 i *> 1.1.1.2 0 200 400 i *> 10.10.0.0/24 2.2.2.3 0 200 400 i * 1.1.1.2 12345 0 200 400 i *m 10.10.1.0/24 2.2.2.3 0 200 400 i *> 1.1.1.2 0 200 400 i *m 172.21.0.0/24 1.1.1.2 0 0 200 i *> 2.2.2.3 0 200 i
The command bgp deterministic-med ensures that an accurate MED comparison is made across all routes received from the same autonomous system (AS).
Cisco recommends using the bgp always-compare-med command, which ensures BGP MED decisions are always deterministic.