BGP聯盟是由單一AS細分多個子AS的方法，其目的包含解決水平分割在距離向量網路的缺點、也能減緩路由器維護BGP會談的負擔。
對於相同AS的路由器而言，iBGP必須有完全互連的連接才能正常運作，為了減緩路由器的負擔，主要有兩種方法：BGP路由器反射、BGP聯盟。
BGP路由反射我已在另一文章發表，在這裡將會把重點放在BGP聯盟。
請見拓樸：
R1、R3、R4是企業網路並運作EIGRP與DMVPN，R1是DMVPN hub、R3與R4是DMVPN spoke，R2是ISP。R3與R4的loopback1是模擬各自的區域網路。
R1、R3、R4的BGP對外AS號碼為134，各自的子AS號碼分別是1、3、4。R1、R3、R4的DMVPN網路為134.0.0.0 / 8。R1、R3、R4的EIGRP AS號碼為134。R5是位於BGP AS號碼為5的路由器。
完整配置請見下面
關鍵命令是bgp mode下的bgp confederation identifier與bgp confederation peers，identifier就是配置對外AS號碼，而peers就是子AS號碼。
在網路完成收斂後，BGP的路由傳播會像這樣：
請注意紅底線部分，R1透過BGP接收網路40.0.0.0 / 8的路徑為(4) i，R3則是(1 4) i，這看起來再普通不過，但是我們來看R5：
R5透過BGP接收網路40.0.0.0 / 8的路徑為134 i，即對R5而言，R1、R3、R4都來自同一個AS(134)，並且R5並不知道134內還有1、3和4。
然而R1、R3、R4可以用類似eBGP的維運方式達到iBGP完全互連的要求，以保證減緩路由器的負擔、解決水平分割的缺點。這就是BGP confederation。
BGP confederation is a method that single AS to multi sub-AS. The purpose include resolve disadvantage for split-horizon in distance-vector, and reduce router loading for maintenance BGP session.
For routers at same AS, working normally that iBGP are must established fully mesh connection. To reduce loading of router, there are 2 methods: BGP route reflector, BGP confederation.
I had writing BGP route reflector in other. Here, I focus in BGP confederation.
You can see topology:
[topology]
BGP confederation is a method that single AS to multi sub-AS. The purpose include resolve disadvantage for split-horizon, and reduce router loading for maintenance BGP session.
For routers at same AS, working normally that iBGP are must established fully mesh connection. To reduce loading of router, there are 2 methods: BGP route reflector, BGP confederation.
I had writing BGP route reflector in other. Here, I focus in BGP confederation.
You can see topology
[topology]
The R1, R3, and R4 are enterprise network and running EIGRP, DMVPN. R1 is DMVPN hub. R3 and R4 are DMVPN spokes. R2 is ISP. R3 and R4 has loopback 1 for simulate LAN self.
The R1, R3, and R4 are BGP AS 134(it is external AS). Their each sub-AS number are 1, 3, and 4. And their DMVPN network is 134.0.0.0 / 8. And their EIGRP AS is 134.
R5 is router located BGP AS 5.
The complete configuration(R1, R2, R3, R4 & R5) in below.
[configuration]
The important commands are 「bgp confederation identifier」 and 「bgp confederation peers」 in bgp mode.
The parameter of 「identifier」 is to set external AS number, and parameter of 「peers」 is to set sub-AS number.
After network has finished converged, The BGP propagate routings are like it:
[R1]
[R3]
Please notice red underline. R1 received network 40.0.0.0/8 by BGP that path is (4). And R3 is (1 4) i. It look very ordinary, but we to see R5:
[R5]
The R5 received network 40.0.0.0 / 8 by BGP that path is 134 i. For R5, there are R1, R3 and R4 from same AS(134). And R5 doesn't know AS 134 includes 1, 3 and 4
However, R1, R3 and R4 can to simulate eBGP-like maintenance method that achieve requirement for iBGP full mesh. It ensure reduce loading for router, and resolve disadvantage for split-horizon.
This is BGP confederation.
留言列表
