Route Summarization
Summarizing prefixes conserves router resource(s) and accelerates best path calculation by reducing the size of the table. Summarization also provides the benefit(s) of stability by reducing routing churn by hiding route flaps from downstream routers. Although most ISPs do not accept prefixes larger than /24 for IPv4 (/25-/32), the Internet, at the time of this writing, still has more than 600,000 routes and continues to grow toward a million routes. Route summarization is required to reduce the size of the BGP table for Internet routers.
BGP route summarization on EBGP routers for nontransitive ASs reduce route computation on routers in the core of the nontransitive AS. In Figure 1-12, R3 summarizes all the EBGP routes received from AS65100 and AS65200 to reduce route computation on R4 during link flaps. In the event of a link flap on the 10.1.13.0/24 network, R3 removes all AS65100 routes learned directly from R1 and identifies the same networks via R2 with a different (longer AS_PATH). R4 processes the same changes that R3 processes and is a waste of CPU cycles because R4 receives connectivity only from R3. If R3 summarized the network range, instead of running the best-path algorithm against multiple routes, the best-path algorithm would execute only once.
)
Figure 1-12 BGP Route Summarization
The two techniques for BGP summarization are the following:
Static: Create a static route to Null 0 for the prefix, and then advertise the network via a network statement. The downfall to this technique is that the summary route will always be advertised even if the networks are not available.
Dynamic: Configure an aggregation network range. When viable routes that match the network range enter the BGP table, an aggregate route is created. On the originating router, the aggregated prefix sets the next-hop to Null 0. The route to Null 0 is automatically created by BGP as a loop-prevention mechanism.
In both methods of route aggregation, a new network prefix with a shorter prefix length is advertised into BGP. Because the aggregated prefix is a new route, the summarizing router is the originator for the new aggregate route.
Aggregate-Address
Dynamic route summarization is accomplished with the BGP address-family configuration commands identified in Table 1-6.
Table 1-6 BGP Route-Aggregation Commands
OS |
Command |
IOS |
aggregate-address network subnet-mask [summary-only | suppress-map route-map-name] [as-set] [advertise-map route-map-name] |
IOS XR |
aggregate-address network/prefix-length [summary-only | route-policy route-policy-name] [as-set] [advertise-map route-policy-name] |
NX-OS |
aggregate-address {network subnet-mask network/prefix-length}[summary- only | suppress-map route-map-name] [as-set] [advertise-map route-map-name] |
The aggregate-address command advertises the aggregated route in addition to the original networks. Using the optional no-summary keyword suppresses the networks in the summarized network range. BGP considers aggregated addresses as local routes.
Flexible Route Suppression
Some traffic engineering designs require “leaking” routes, which is the advertisement of a subset of more specific routes in addition to performing the summary. Leaking routes can be done at the process by explicitly stating the prefixes to suppress, or on a neighbor level by indicating which prefixes should not be suppressed.
Selective Prefix Suppression
Selective prefix suppression explicitly lists the networks that should not be advertised along with the summary route to neighbor routers.
IOS and NX-OS uses a suppress-map, which uses the keyword suppress-map route-map-name instead of using the no-summary keyword. In the referenced route-map, only the prefixes that should be suppressed are permitted. IOS XR routers use the keyword route-policy route-policy-name in lieu of the no-summary keyword. In the route policy, the action command suppress is used after conditionally matching the prefixes that should be suppressed.
Leaking Suppressed Routes
The summary-only keyword suppresses all the more specific routes of an aggregate address from being advertised. After a route is suppressed, it is still possible to advertise the suppressed route to a specific neighbor.
IOS devices use an unsuppress-map with the BGP neighbor address-family configuration command neighbor ip-address unsuppress-map route-map-name. In the referenced route-map, only the prefixes that should be leaked are permitted. IOS XR routers use an outbound route policy with the action command unsuppress to indicate which prefixes should be leaked.
Atomic Aggregate
Aggregated routes act like new BGP routes with a shorter prefix length. When a BGP router summarizes a route, it does not advertise the AS path information from before the aggregation. BGP path attributes such as AS-Path, MED, and BGP communities are not included in the new BGP advertisement. The Atomic Aggregate attribute indicates that a loss of path information has occurred.
For example:
R1 and R2 are advertising the 172.16.1.0/24 and 172.16.2.0/24 networks.
R3 is aggregating the routes into the 172.16.0.0/22 network range, which is advertised to all of R3’s peers
Example 1-14 displays R3’s BGP table. R1’s BGP prefix 172.16.1.0/24 advertised to R3. Notice the AS-Path of 65100 and BGP Community of 100:100.
Example 1-14 172.16.1.0/24 BGP Path Information
R3-IOS# show bgp ipv4 unicast 172.16.1.0
BGP routing table entry for 172.16.1.0/24, version 13
Paths: (1 available, best #1, table default, Advertisements suppressed by an aggre-
gate.)
Not advertised to any peer
Refresh Epoch 1
65100
10.1.13.1 from 10.1.13.1 (192.168.1.1)
Origin IGP, metric 0, localpref 100, valid, external, best
Community: 100:100
R3’s aggregate route (summary) does not include the BGP communities (including AS-Path history) for the routes in the summarization range. R3 advertises the aggregate route to R1 and R2, and those routers install the 172.16.0.0/22 summary route because their AS-Path is not listed in the AS-Path attribute and passes the AS-Path loop check.
Example 1-15 displays the BGP path information for the 172.16.0.0/22 summary network on R1. The AS-Path of the aggregated route displays only the aggregating router, but does not include the AS-Path of the routes being summarized (AS65100 or AS65200), nor is the BGP community included in the routes being summarized. The BGP path information indicates that this is an aggregated prefix and was aggregated by R3 (192.168.3.3). The Atomic-Aggregate in the route indicates a loss of information occurred during aggregation on the aggregating router.
Example 1-15 172.16.0.0/22 BGP Path Information
R1-IOS# show bgp ipv4 unicast 172.16.0.0
BGP routing table entry for 172.16.0.0/21, version 5
Paths: (1 available, best #1, table default)
Not advertised to any peer
Refresh Epoch 1
300, (aggregated by 300 192.168.3.3)
10.1.13.3 from 10.1.13.3 (192.168.3.3)
Origin IGP, metric 0, localpref 100, valid, external, atomic-aggregate, best
Route Aggregation with AS_SET
To keep the BGP path information history, the optional as-set keyword may be used with the aggregate-address command. As the router generates the aggregate route, BGP attributes from the summarized routes are copied over to it. The AS-Path settings from the original prefixes are stored in the AS_SET portion of the AS-Path. (The AS_SET is displayed within brackets, and counts only as one hop, even if multiple ASs are listed.)
Route Aggregation with Selective Advertisement of AS-SET
Using the AS-SET feature with network aggregation combines all the attributes of the original prefixes into the aggregated prefixes. This might cause issues with your routing policy. For example, if one of the prefixes contains the No-Export BGP community, the aggregate address will not be exported. To resolve these types of problems, selectively choose the routes that the path attributes will copy to the aggregate route. The use of the advertise-map option allows for conditionally matching and denying attributes that should be permitted or denied in the aggregated route.
Default Route Advertisement
Advertising a default route into the BGP table requires the default route to exist in the RIB and the BGP configuration command default-information originate to be used. The redistribution of a default route or use of a network 0.0.0.0/0 does not work without the default-information originate command.
Default Route Advertisement per Neighbor
Some network topologies restrict the size of the BGP advertisements to a neighbor because the remote router does not have enough processing power or memory for the full BGP routing table. Connectivity is still required, so the peering routers only advertise the default route to the remote router.
A default route is advertised to a BGP peer with the BGP address-family configuration command neighbor ip-address default-originate for IOS nodes or with the BGP neighbor address-family configuration command default-originate for IOS XR and NX-OS devices. Default route advertisement to a specific neighbor does not require a default route to be present in the RIB or BGP Loc-RIB table.