含MPLS設計的Cisco IOS-XR BGP
目錄
簡介
本檔案介紹幾個案例，其中針對Cisco IOS®-XR中的多重通訊協定標籤交換(MPLS)和邊界閘道通訊協定(BGP)的組合，具有特殊的行為和組態。
1. Inter-AS MPLS VPN選項B & C需要靜態路由
此圖顯示Inter-AS選項B設定。
圖1.
提供商邊緣(PE)路由器PE1具有用於VRF字首10.200.1.2/32的路由，但它未解析。
RP/0/0/CPU0:PE1#show cef vrf one 10.200.1.2
10.200.1.2/32, version 3, internal 0x5000001 0x0 (ptr 0xa140be74) [1], 0x0 (0x0), 0x208 (0xa14a7118)
Updated Apr 7 14:36:45.628
Prefix Len 32, traffic index 0, precedence n/a, priority 3
via 10.3.1.4/32, 0 dependencies, recursive [flags 0x6000]
path-idx 0 NHID 0x0 [0xa0d87468 0x0]
recursion-via-/32
next hop VRF - 'default', table - 0xe0000000
unresolved
labels imposed {24004}
PE1沒有10.3.1.4/32的路由，但有10.3.1.0/24的路由。
RP/0/0/CPU0:PE1#show route 10.3.1.4
Routing entry for 10.3.1.0/24
Known via "ospf 1", distance 110, metric 3, type intra area
Installed Apr 7 14:07:01.140 for 00:32:48
Routing Descriptor Blocks
10.1.1.2, from 10.100.1.3, via GigabitEthernet0/0/0/0
Route metric is 3
No advertising protos.
自治系統邊界路由(ASBR)上必須有下一跳的靜態路由。您必須在每個ASBR上配置此靜態路由，並將其重新分配到內部網關協定(IGP)中。
router static
address-family ipv4 unicast
10.3.1.4/32 GigabitEthernet0/0/0/1
!
!
router ospf 1
redistribute static
路由現在已解析。
RP/0/0/CPU0:PE1#show cef vrf one 10.200.1.2
10.200.1.2/32, version 3, internal 0x5000001 0x0 (ptr 0xa140be74) [1], 0x0 (0x0), 0x208 (0xa14a7118)
Updated Apr 7 14:36:45.628
Prefix Len 32, traffic index 0, precedence n/a, priority 3
via 10.3.1.4/32, 3 dependencies, recursive [flags 0x6000]
path-idx 0 NHID 0x0 [0xa150f9f4 0x0]
recursion-via-/32
next hop VRF - 'default', table - 0xe0000000
next hop 10.3.1.4/32 via 24005/0/21
next hop 10.1.1.2/32 Gi0/0/0/0 labels imposed {24003 24004}
ASBR1為VPNv4/6字首安裝指向ASBR2的POP傳出標籤：
RP/0/0/CPU0:ASBR1#show mpls forwarding prefix 10.3.1.4/32
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
24005 Pop 10.3.1.4/32 Gi0/0/0/1 10.3.1.4 2506
即使ASBR上使用指向iBGP鄰居的下一跳自身，如果ASBR上未配置靜態路由，ASBR之間的標籤轉發也將中斷。
在ASBR1上使用指向PE1的下一跳self，並且沒有靜態路由：
RP/0/0/CPU0:ASBR1#show mpls forwarding labels 24006 detail
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
24006 24004 2:2:10.200.1.2/32 10.3.1.4 0
Updated: Apr 7 14:49:58.190
Path Flags: 0x6000 [ ]
Label Stack (Top -> Bottom): { }
MAC/Encaps: 0/0, MTU: 0
Packets Switched: 0
請注意，「傳出介面」列中缺少傳出介面。Inter-AS選項B和C的ASBR上需要靜態路由。
2.在ASBR上保留Inter-AS選項B的路由目標
需要命令來確保ASBR儲存/保留vpnv4/6路由，然後通告它們。如果沒有此命令，如果ASBR上沒有配置匯入路由的任何路由目標的本地VRF，或者它不是地址系列vpnv4/6的路由反射器(RR)，則ASBR不會儲存路由。
router bgp 1
address-family ipv4 unicast
!
address-family vpnv4 unicast
retain route-target all
!
3. ASBR不通告IPv4標籤的單點傳播地址
Inter-AS選項C或無縫MPLS（統一MPLS）網路中需要IPv4 labeled-unicast。這是因為預設情況下會標籤vpnv4/6字首，但IPv4(IPv6)單播的情況並非如此。如果情況並非如此，則標籤交換路徑(LSP)端對端會中斷，流量會端對端失敗。
檢視圖2，其中顯示Inter-AS Option C設定。
圖2.
P1和P2路由器也是vpnv4的自治系統(AS)中的路由反射器。
標籤單播(LU)用於將環回字首從一個AS傳輸到另一個AS。
ASBR1已配置此地址系列，但其中沒有路由：
RP/0/0/CPU0:ASBR1#show bgp ipv4 labeled-unicast
RP/0/0/CPU0:ASBR1#
RP/0/0/CPU0:ASBR1#show bgp ipv4 labeled-unicast summary
BGP router identifier 10.100.1.3, local AS number 1
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0xe0000000 RD version: 41
BGP main routing table version 41
BGP NSR Initial initsync version 2 (Reached)
BGP NSR/ISSU Sync-Group versions 0/0
BGP scan interval 60 secs
BGP is operating in STANDALONE mode.
Process RcvTblVer bRIB/RIB LabelVer ImportVer SendTblVer StandbyVer
Speaker 41 41 41 41 41 0
Neighbor Spk AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down St/PfxRcd
10.3.1.4 0 2 150 151 41 0 0 00:06:29 0
10.100.1.2 0 1 52 52 41 0 0 00:06:42 0
原因是ASBR必須使用以下命令，以便可以為每個路由分配多協定標籤交換(MPLS)標籤，然後通告路由。
RP/0/0/CPU0:ASBR1#show run router bgp
router bgp 1
address-family ipv4 unicast
redistribute ospf 1
allocate-label all
!
此命令的結果如下：
RP/0/0/CPU0:ASBR1#show bgp ipv4 labeled-unicast
BGP router identifier 10.100.1.3, local AS number 1
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0xe0000000 RD version: 52
BGP main routing table version 52
BGP NSR Initial initsync version 2 (Reached)
BGP NSR/ISSU Sync-Group versions 0/0
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
i - internal, r RIB-failure, S stale, N Nexthop-discard
Origin codes: i - IGP, e - EGP, ? - incomplete
Network Next Hop Metric LocPrf Weight Path
*> 10.1.1.0/24 10.1.2.2 2 32768 ?
*> 10.1.2.0/24 0.0.0.0 0 32768 ?
*> 10.2.1.0/24 10.3.1.4 0 0 2 ?
*> 10.2.2.0/24 10.3.1.4 2 0 2 ?
*> 10.3.1.0/24 0.0.0.0 0 32768 ?
* 10.3.1.4 0 0 2 ?
*> 10.100.1.1/32 10.1.2.2 3 32768 ?
*> 10.100.1.2/32 10.1.2.2 2 32768 ?
*> 10.100.1.3/32 0.0.0.0 0 32768 ?
*> 10.100.1.4/32 10.3.1.4 0 0 2 ?
*> 10.100.1.5/32 10.3.1.4 2 0 2 ?
*> 10.100.1.6/32 10.3.1.4 3 0 2 ?
Processed 11 prefixes, 12 paths
RP/0/0/CPU0:ASBR1#show bgp ipv4 labeled-unicast 10.100.1.6/32
BGP routing table entry for 10.100.1.6/32
Versions:
Process bRIB/RIB SendTblVer
Speaker 48 48
Local Label: 24008
Last Modified: Apr 7 16:20:04.509 for 00:00:49
Paths: (1 available, best #1)
Advertised to peers (in unique update groups):
10.100.1.2
Path #1: Received by speaker 0
Advertised to peers (in unique update groups):
10.100.1.2
2
10.3.1.4 from 10.3.1.4 (10.100.1.4)
Received Label 24002
Origin incomplete, metric 3, localpref 100, valid, external, best, group-best
Received Path ID 0, Local Path ID 1, version 48
Origin-AS validity: not-found
所以，簡而言之：
- 必須配置標籤分配（在所有ASBR和PE路由器上）
- 必須為BGP鄰居配置地址系列ipv4 labeled-unicast
4. eBGP-eBGP（背對背的eBGP）Vpnv4需要虛擬iBGP鄰居
看圖3。
圖3.
連續有三個ASBR。ASBR3運行eBGP vpnv4單播到ASBR1和ASBR2。
RP/0/0/CPU0:ASBR3#show bgp vpnv4 unicast
BGP router identifier 10.100.1.7, local AS number 3
BGP generic scan interval 60 secs
Non-stop routing is enabled
BGP table state: Active
Table ID: 0x0 RD version: 0
BGP main routing table version 3
BGP NSR Initial initsync version 2 (Reached)
BGP NSR/ISSU Sync-Group versions 0/0
BGP scan interval 60 secs
Status codes: s suppressed, d damped, h history, * valid, > best
i - internal, r RIB-failure, S stale, N Nexthop-discard
Origin codes: i - IGP, e - EGP, ? - incomplete
Network Next Hop Metric LocPrf Weight Path
Route Distinguisher: 1:1
*> 10.200.1.1/32 10.4.1.3 0 1 ?
Route Distinguisher: 2:2
*> 10.200.1.2/32 10.4.2.4 0 2 ?
Processed 2 prefixes, 2 paths
RP/0/0/CPU0:ASBR3#show bgp vpnv4 unicast rd 1:1 10.200.1.1/32
BGP routing table entry for 10.200.1.1/32, Route Distinguisher: 1:1
Versions:
Process bRIB/RIB SendTblVer
Speaker 2 2
Last Modified: Apr 7 18:45:21.510 for 00:03:30
Paths: (1 available, best #1)
Not advertised to any peer
Path #1: Received by speaker 0
Not advertised to any peer
1
10.4.1.3 from 10.4.1.3 (10.100.1.3)
Received Label 24009
Origin incomplete, localpref 100, valid, external, best, group-best, import-candidate, not-in-vrf
Received Path ID 0, Local Path ID 1, version 2
Extended community: RT:1:1
從ASBR3通告vpnv4路由時出現問題：ASBR3不會向外通告外部vpnv4路由。
解決方案是在ASBR3上配置虛擬iBGP鄰居並啟用next-hop-self:虛構iBGP鄰居無需啟動。
router bgp 3
address-family vpnv4 unicast
retain route-target all
!
neighbor 10.4.1.3
remote-as 1 address-family vpnv4 unicast
route-policy PASS in
route-policy PASS out
!
!
neighbor 10.4.2.4
remote-as 2
address-family vpnv4 unicast
route-policy PASS in
route-policy PASS out
!
!
neighbor 10.99.99.99
remote-as 3
description dummy-iBGP neighbor for back-to-back eBGP vpnv4
update-source Loopback0
address-family vpnv4 unicast
next-hop-self
!
!
!
結果是vpnv4路由現在被通告：
RP/0/0/CPU0:ASBR3#show bgp vpnv4 unicast rd 1:1 10.200.1.1/32
BGP routing table entry for 10.200.1.1/32, Route Distinguisher: 1:1
Versions:
Process bRIB/RIB SendTblVer
Speaker 12 12
Local Label: 24002
Last Modified: Apr 7 18:58:04.510 for 00:01:46
Paths: (1 available, best #1)
Advertised to update-groups (with more than one peer):
0.2
Path #1: Received by speaker 0
Advertised to update-groups (with more than one peer):
0.2
1
10.4.1.3 from 10.4.1.3 (10.100.1.3)
Received Label 24009
Origin incomplete, localpref 100, valid, external, best, group-best, import-candidate, not-in-vrf
Received Path ID 0, Local Path ID 1, version 12
Extended community: RT:1:1
5. Inter-AS選項C — 首選BGP標籤而非LDP標籤
請參閱此映像，檢視通過多條鏈路連線的兩個ASBR的設定。為了讓此起作用，ASBR之間的eBGP ipv4 LU會話必須是多跳，因為它們之間具有並行鏈路。
圖4.
這是Inter-AS選項C。路由器P1和P2也是vpnv4的路由反射器。
PE路由器和ASBR之間有IPv4標籤的單播。ASBR通過多條鏈路直接連線。
在ASBR上，您會看到：
router bgp 1
…
neighbor 10.100.1.4
remote-as 2
ebgp-multihop 2
update-source Loopback0
address-family ipv4 labeled-unicast
route-policy PASS in
route-policy PASS out
ASBR之間不需要標籤分發協定(LDP)。BGP將處理ASBR之間鏈路上的MPLS轉發。
RP/0/0/CPU0:ASBR1#show mpls interfaces
Interface LDP Tunnel Static Enabled
-------------------------- -------- -------- -------- --------
GigabitEthernet0/0/0/0 Yes No No Yes
GigabitEthernet0/0/0/1 No No No Yes
GigabitEthernet0/0/0/2 No No No Yes
GigabitEthernet0/0/0/3 No No No Yes
GigabitEthernet0/0/0/4 No No No Yes
目前為止一切順利。問題出在此圖所示的情境。
圖5.
這是Inter-AS選項C。路由器P1和P2也是vpnv4的路由反射器。
PE路由器和ASBR之間有IPv4標籤的單播。ASBR1和ASBR2未直接連線。它們通過運行IGP和LDP的網路進行多跳連線。在映像5中，此中間網路由路由器ASBR3表示，該路由器使用ASBR1和ASBR2運行IGP和LDP。
在ASBR上使用eBGP多躍點時，出現問題。每個AS中的RR之間的BGP會話甚至不啟動。
RP/0/0/CPU0:P1#show cef 10.100.1.5
10.100.1.5/32, version 263, internal 0x1000001 0x0 (ptr 0xa13bde74) [1], 0x0 (0xa1389560), 0xa28 (0xa14a72a8)
Updated Apr 8 09:38:02.551
local adjacency 10.1.2.3
Prefix Len 32, traffic index 0, precedence n/a, priority 3
via 10.1.2.3/32, GigabitEthernet0/0/0/1, 5 dependencies, weight 0, class 0 [flags 0x0]
path-idx 0 NHID 0x0 [0xa0e8b2a4 0x0]
next hop 10.1.2.3/32
local adjacency
local label 24004 labels imposed {24007}
為了從P1、AS 1中的RR到P2、AS 2中的RR，傳出標籤為24007。在ASBR1上，此標籤交換為標籤24000。
RP/0/0/CPU0:ASBR1#show mpls forwarding labels 24007
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
24007 24000 10.100.1.5/32 10.100.1.4 1404
RP/0/0/CPU0:ASBR1#show cef 10.100.1.5
10.100.1.5/32, version 155, internal 0x5000001 0x0 (ptr 0xa13be174) [1], 0x0 (0xa138965c), 0xa08 (0xa14a72d0)
Updated Apr 8 10:02:38.101
Prefix Len 32, traffic index 0, precedence n/a, priority 4
via 10.100.1.4/32, 5 dependencies, recursive, bgp-ext [flags 0x6020]
path-idx 0 NHID 0x0 [0xa150f874 0x0]
recursion-via-/32
next hop 10.100.1.4/32 via 24004/0/21
local label 24007
next hop 10.4.1.7/32 Gi0/0/0/4 labels imposed {ImplNull 24000}
標籤24000是BGP LU從ASBR2在ASBR1上接收的標籤。
RP/0/0/CPU0:ASBR1#show bgp ipv4 labeled-unicast 10.100.1.5
BGP routing table entry for 10.100.1.5/32
Versions:
Process bRIB/RIB SendTblVer
Speaker 76 76
Local Label: 24007
Last Modified: Apr 8 09:37:57.509 for 00:04:05
Paths: (1 available, best #1)
Advertised to update-groups (with more than one peer):
0.3
Advertised to peers (in unique update groups):
10.100.1.1 10.100.1.2
Path #1: Received by speaker 0
Advertised to update-groups (with more than one peer):
0.3
Advertised to peers (in unique update groups):
10.100.1.1 10.100.1.2
2
10.100.1.4 from 10.100.1.4 (10.100.1.4)
Received Label 24000
Origin incomplete, metric 2, localpref 100, valid, external, best, group-best
Received Path ID 0, Local Path ID 1, version 76
Origin-AS validity: not-found
但是，介於兩者之間的ASBR路由器不運行BGP，因此無法轉發它收到的帶有此標籤的資料包，因為它沒有分配標籤24000。應該用於使資料包到達10.100.1.5的標籤是來自LDP的標籤：
RP/0/0/CPU0:ASBR1#show route 10.100.1.5/32
Routing entry for 10.100.1.5/32
Known via "bgp 1", distance 20, metric 2, [ei]-bgp, labeled unicast (3107)
Tag 2, type external
Installed Apr 8 10:02:38.082 for 01:24:37
Routing Descriptor Blocks
10.100.1.4, from 10.100.1.4, BGP external
Route metric is 2
No advertising protos.
這會遞迴到下一跳10.100.1.4，即ASBR2的環回。
應使用LDP從ASBR3接收的標籤，但不應使用。
新增的標籤堆疊為{ImplNull 24000}，而不是{24002 24000}。
RP/0/0/CPU0:ASBR1#show mpls ldp bindings 10.100.1.4/32
10.100.1.4/32, rev 146
Local binding: label: 24004
Remote bindings: (2 peers)
Peer Label
----------------- ---------
10.100.1.2:0 24003
10.100.1.7:0 24002
ASBR1應該從ASBR3路由器24002收到的LDP標籤強制。若要停用BGP MPLS轉送，請將mpls關鍵字新增到eBGP multi-hop命令中。
ASBR1:
router bgp 1
…
neighbor 10.100.1.4
remote-as 2
ebgp-multihop 2 mpls
update-source Loopback0
address-family ipv4 labeled-unicast
route-policy PASS in
route-policy PASS out
!
ASBR1現在具有正確的標籤重寫：
RP/0/0/CPU0:ASBR1#show cef 10.100.1.5
10.100.1.5/32, version 155, internal 0x5000001 0x0 (ptr 0xa13be174) [1], 0x0 (0xa138965c), 0xa08 (0xa14a72d0)
Updated Apr 8 10:02:38.102
Prefix Len 32, traffic index 0, precedence n/a, priority 4
via 10.100.1.4/32, 5 dependencies, recursive, bgp-ext [flags 0x6020]
path-idx 0 NHID 0x0 [0xa150f874 0x0]
recursion-via-/32
next hop 10.100.1.4/32 via 24004/0/21
local label 24007
next hop 10.4.1.7/32 Gi0/0/0/4 labels imposed {24002 24000}
在命令引用中：
在ebgp-multihop命令中使用mpls選項會阻止BGP在對等介面上啟用MPLS，還會阻止為從對等裝置獲取的下一跳地址分配隱式 — NULL重寫標籤。這在已通過BGP標籤 — 單播或LDP學習到MPLS轉發標籤到下一跳的某些場景中很有用。
換句話說，在IOS-XR中，當BGP提供向LFIB分配標籤時，它將優先於LDP。在ASBR路由器之間有多個躍點的Inter-AS選項C就是這樣一種場景。
6. Inter-AS選項B — 首選BGP標籤而非LDP標籤
圖6.
這是Inter-AS選項B。但是，兩個ASBR之間存在多個並行鏈路。ASBR之間有RFC3107（交換IPv4路由和MPLS標籤），而不是使用IGP和LDP。
為了在ASBR1和ASBR2的環回介面之間啟動eBGP多跳會話，兩個ASBR之間需要eBGP LU。ASBR之間有兩條鏈路，因此需要兩個eBGP LU會話。地址系列IPv4需要命令allocate-label。
router bgp 65001
address-family ipv4 unicast
network 10.100.1.3/32
allocate-label all
!
neighbor 10.3.1.4
remote-as 65002
address-family ipv4 labeled-unicast
route-policy pass in
route-policy pass out
!
!
neighbor 10.3.2.4
remote-as 65002
address-family ipv4 labeled-unicast
route-policy pass in
route-policy pass out
!
!
仍然需要來自第1部分的靜態路由：
router static
address-family ipv4 unicast
10.3.1.4/32 GigabitEthernet0/0/0/1
10.3.2.4/32 GigabitEthernet0/0/0/2
!
!
ASBR之間的eBGP vpnv4會話：
router bgp 65001
address-family ipv4 unicast
network 10.100.1.3/32
allocate-label all
!
address-family vpnv4 unicast
retain route-target all
!
neighbor 10.100.1.4
remote-as 65002
ebgp-multihop 255
update-source Loopback0
address-family vpnv4 unicast
route-policy pass in
route-policy pass out
!
!
請注意，此處不需要mpls關鍵字，如第5節所述。此外，如果為iBGP vpnv4會話配置了next-hop-self，則不需要在PE和ASBR之間執行iBGP LU會話。ASBR2為10.100.1.4/32通告的標籤為標籤3:
RP/0/0/CPU0:ASBR1#show bgp ipv4 labeled-unicast 10.100.1.4/32
Fri Jun 2 11:50:16.178 UTC
BGP routing table entry for 10.100.1.4/32
Versions:
Process bRIB/RIB SendTblVer
Speaker 8 8
Local Label: 24005
Last Modified: Jun 2 11:48:39.920 for 00:01:36
Paths: (4 available, best #1)
Advertised to update-groups (with more than one peer):
0.3
Advertised to peers (in unique update groups):
10.100.1.7
Path #1: Received by speaker 0
Advertised to update-groups (with more than one peer):
0.3
Advertised to peers (in unique update groups):
10.100.1.7
65002
10.3.1.4 from 10.3.1.4 (10.100.1.4)
Received Label 3
Origin IGP, metric 0, localpref 100, valid, external, best, group-best
Received Path ID 0, Local Path ID 1, version 8
Origin-AS validity: not-found
Path #2: Received by speaker 0
Not advertised to any peer
65002
10.3.2.4 from 10.3.2.4 (10.100.1.4)
Received Label 3
Origin IGP, metric 0, localpref 100, valid, external
Received Path ID 0, Local Path ID 0, version 0
Origin-AS validity: not-found
Path #3: Received by speaker 0
Not advertised to any peer
65003 65002
10.3.3.9 from 10.3.3.9 (10.100.1.9)
Received Label 24001
Origin IGP, localpref 100, valid, external, group-best
Received Path ID 0, Local Path ID 0, version 0
Origin-AS validity: not-found
Path #4: Received by speaker 0
Not advertised to any peer
65003 65002
10.3.4.9 from 10.3.4.9 (10.100.1.9)
Received Label 24001
Origin IGP, localpref 100, valid, external
Received Path ID 0, Local Path ID 0, version 0
Origin-AS validity: not-found
RP/0/0/CPU0:ASBR1#show cef 10.100.1.4
Fri Jun 2 11:51:06.994 UTC
10.100.1.4/32, version 254, internal 0x1000001 0x0 (ptr 0xa13be474) [1], 0x0 (0xa13896ec), 0xa20 (0xa14a70f0)
Updated Jun 2 11:48:39.634
local adjacency 10.3.1.4
Prefix Len 32, traffic index 0, precedence n/a, priority 4
via 10.3.1.4/32, GigabitEthernet0/0/0/1, 5 dependencies, weight 0, class 0 [flags 0x0]
path-idx 0 NHID 0x0 [0xa0e8b1fc 0xa0e8b34c]
next hop 10.3.1.4/32
local adjacency
local label 24005 labels imposed {ImplNull}
RP/0/0/CPU0:ASBR1#show mpls forwarding labels 24005
Fri Jun 2 11:51:20.204 UTC
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
24005 Pop 10.100.1.4/32 Gi0/0/0/1 10.3.1.4 610
當ASBR之間存在另一個路徑，並且該路徑使用IGP + LDP或MPLS TE時，eBGP多跳命令需要mpls關鍵字。
圖7.
ASBR1上指向P3的BGP路由策略用於設定非常高的權重，因此來自P3的字首優先於來自ASBR2的字首。
RP/0/0/CPU0:ASBR1#show bgp ipv4 labeled-unicast 10.100.1.4/32
Fri Jun 2 11:57:23.789 UTC
BGP routing table entry for 10.100.1.4/32
Versions:
Process bRIB/RIB SendTblVer
Speaker 9 9
Local Label: 24005
Last Modified: Jun 2 11:51:58.920 for 00:05:24
Paths: (4 available, best #3)
Advertised to update-groups (with more than one peer):
0.3
Advertised to peers (in unique update groups):
10.100.1.7
Path #1: Received by speaker 0
Not advertised to any peer
65002
10.3.1.4 from 10.3.1.4 (10.100.1.4)
Received Label 3
Origin IGP, metric 0, localpref 100, valid, external, group-best
Received Path ID 0, Local Path ID 0, version 0
Origin-AS validity: not-found
Path #2: Received by speaker 0
Not advertised to any peer
65002
10.3.2.4 from 10.3.2.4 (10.100.1.4)
Received Label 3
Origin IGP, metric 0, localpref 100, valid, external
Received Path ID 0, Local Path ID 0, version 0
Origin-AS validity: not-found
Path #3: Received by speaker 0
Advertised to update-groups (with more than one peer):
0.3
Advertised to peers (in unique update groups):
10.100.1.7
65003 65002
10.3.3.9 from 10.3.3.9 (10.100.1.9)
Received Label 24001
Origin IGP, localpref 100, weight 65535, valid, external, best, group-best
Received Path ID 0, Local Path ID 1, version 9
Origin-AS validity: not-found
Path #4: Received by speaker 0
Not advertised to any peer
65003 65002
10.3.4.9 from 10.3.4.9 (10.100.1.9)
Received Label 24001
Origin IGP, localpref 100, valid, external
Received Path ID 0, Local Path ID 0, version 0
Origin-AS validity: not-found
ASBR1現在應使用label 24001作為10.100.1.4/32的傳出標籤。它不：
RP/0/0/CPU0:ASBR1#show cef 10.100.1.4
Fri Jun 2 11:59:46.519 UTC
10.100.1.4/32, version 255, internal 0x1000001 0x0 (ptr 0xa13be474) [1], 0x0 (0xa13896ec), 0xa20 (0xa14a7140)
Updated Jun 2 11:51:58.741
local adjacency 10.3.3.9
Prefix Len 32, traffic index 0, precedence n/a, priority 4
via 10.3.3.9/32, GigabitEthernet0/0/0/3, 7 dependencies, weight 0, class 0 [flags 0x0]
path-idx 0 NHID 0x0 [0xa0e8b544 0xa0e8b5ec]
next hop 10.3.3.9/32
local adjacency
local label 24005 labels imposed {ImplNull}
解決方案與第5部分相同：將mpls關鍵字用於eBGP多跳命令。
RP/0/0/CPU0:ASBR1# conf t
Fri Jun 2 13:56:45.618 UTC
RP/0/0/CPU0:ASBR1(config)#router bgp 65001
RP/0/0/CPU0:ASBR1(config-bgp)# neighbor 10.100.1.4
RP/0/0/CPU0:ASBR1(config-bgp-nbr)#ebgp-multihop 255 mpls
RP/0/0/CPU0:ASBR1(config-bgp-nbr)#commit
ASBR1現在將標籤24001用作10.100.1.4/32的傳出標籤。
RP/0/0/CPU0:ASBR1#show cef 10.100.1.4
Fri Jun 2 13:58:13.402 UTC
10.100.1.4/32, version 200, internal 0x5000001 0x0 (ptr 0xa13be474) [1], 0x0 (0xa13895cc), 0xa08 (0xa14a71b8)
Updated Jun 2 13:56:59.378
Prefix Len 32, traffic index 0, precedence n/a, priority 15
via 10.3.3.9/32, 3 dependencies, recursive, bgp-ext [flags 0x6020]
path-idx 0 NHID 0x0 [0xa15102f4 0x0]
recursion-via-/32
next hop 10.3.3.9/32 via 24014/0/21
local label 24005
next hop 10.3.3.9/32 Gi0/0/0/3 labels imposed {ImplNull 24001}
ASBR1推送此額外標籤。從PE1到PE2的虛擬路由和轉發(VRF)中的traceroute顯示推送的額外標籤。
RP/0/0/CPU0:PE1#trace vrf one 10.99.1.2
Fri Jun 2 13:49:38.959 UTC
Type escape sequence to abort.
Tracing the route to 10.99.1.2
1 10.1.1.5 [MPLS: Labels 24002/24012 Exp 0] 29 msec 39 msec 39 msec
2 10.1.2.3 [MPLS: Label 24012 Exp 0] 29 msec 29 msec 39 msec
3 10.3.1.4 [MPLS: Label 24007 Exp 0] 39 msec 39 msec 39 msec
4 10.2.1.6 [MPLS: Labels 24001/24005 Exp 0] 39 msec 39 msec 29 msec
5 10.2.2.2 39 msec * 239 msec
ASBR1和P3之間使用IGP和LDP，ASBR2和P3之間使用IGP和LDP。在這些路由器之間使用MPLS流量工程(TE)時，存在相同的問題和解決方案。
從ASBR1到P3沒有LDP，但有MPLS TE。
在eBGP multihop命令上沒有mpls關鍵字，相同的問題會再次出現：
轉發到10.100.1.4的資料包不會獲得BGP LU標24000被推。
RP/0/0/CPU0:ASBR1#show cef 10.100.1.4
Tue Jun 6 10:36:56.528 UTC
10.100.1.4/32, version 50, internal 0x1000001 0x0 (ptr 0xa12cc1fc) [1], 0x0 (0xa12b18c0), 0xa20 (0xa14a7258)
Updated Jun 6 10:36:32.930
Prefix Len 32, traffic index 0, precedence n/a, priority 4
via 10.3.3.9/32, tunnel-te1, 7 dependencies, weight 0, class 0 [flags 0x0]
path-idx 0 NHID 0x0 [0xa15d58f8 0xa15d5840]
next hop 10.3.3.9/32
local adjacency
local label 24012 labels imposed {ImplNull}
而使用mpls關鍵字時，標籤頁24000會出現：
RP/0/0/CPU0:ASBR1#show cef 10.100.1.4
Tue Jun 6 10:36:03.241 UTC
10.100.1.4/32, version 34, internal 0x5000001 0x0 (ptr 0xa12cc1fc) [1], 0x0 (0xa12b15a8), 0xa08 (0xa14a70f0)
Updated Jun 6 09:39:24.56
Prefix Len 32, traffic index 0, precedence n/a, priority 15
Extensions: context-label:24012
via 10.3.3.9/32, 3 dependencies, recursive, bgp-ext [flags 0x6020]
path-idx 0 NHID 0x0 [0xa150fecc 0x0]
recursion-via-/32
next hop 10.3.3.9/32 via 24011/0/21
local label 24012
next hop 10.3.3.9/32 tt1 labels imposed {ImplNull 24000}
使用mpls關鍵字時，重寫如下所示：
RP/0/0/CPU0:ASBR1#show mpls forwarding labels 24012
Tue Jun 6 10:43:50.559 UTC
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
24012 24000 10.100.1.4/32 tt1 10.3.3.9 0
如果不使用mpls關鍵字，則重寫如下所示：
RP/0/0/CPU0:ASBR1#show mpls forwarding labels 24012
Tue Jun 6 10:45:08.734 UTC
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
24012 Pop 10.100.1.4/32 tt1 10.3.3.9 0
此標籤14012不用於從VRF到VRF或PE到PE的流量，但是如果遇到此標籤，則可能表示標籤轉發例項庫(LFIB)條目錯誤或錯誤。
RP/0/0/CPU0:PE1# trace vrf one 10.99.1.2
Type escape sequence to abort.
Tracing the route to 10.99.1.2
1 10.1.1.5 [MPLS: Labels 24001/24015 Exp 0] 129 msec 229 msec 129 msec
2 10.1.2.3 [MPLS: Label 24015 Exp 0] 219 msec 439 msec 349 msec
3 10.3.3.9 [MPLS: Labels 24000/24011 Exp 0] 169 msec 249 msec 139 msec
4 10.3.5.4 [MPLS: Label 24011 Exp 0] 89 msec 129 msec 109 msec
5 10.2.1.6 [MPLS: Labels 24004/24008 Exp 0] 139 msec 99 msec 139 msec
6 10.2.2.2 129 msec * 219 msec
在eBGP multihop命令上切換關鍵字mpls可能會導致BGP標籤衝突的系統日誌消息：
bgp[1051]: %ROUTING-BGP-4-LABEL_COLLISION : Label 24012 collision: prev: [T: 3 RD:0:0:0 PFX/NHID:10.100.1.4/32] curr: [T: 13 RD:0:0:0 PFX/NHID:10.100.1.4/32]
此消息用於本地標籤24012。
檢查是為了確保BGP擁有的活動標籤不會被BGP重新分配到任何其他內容。此檢查僅適用於每個字首標籤。
此郵件是症狀而非導致本文出現任何問題的原因。
7.使用VPN（或6PE或EVPN）底層的多跳BGP會話
如果存在eBGP多重躍點作業階段，則無法透過vpnv4/6或6PE（使用MPLS的IPv6）或乙太網路虛擬私人網路(EVPN)路由得知下一個躍點位址的路由，除非路由器具有Cisco IOS®-XR 6.3.2或更新版本。請參閱此映像。
圖8.
可能的故障情形：
- 從PE1（VRF內）到PE2（VRF內）的eBGP多跳會話
- 從PE1（VRF內部）到CE2的eBGP多跳會話
這適用：
eBGP多跳會話配置在PE路由器上路由器BGP下的VRF部分下。
從PE1（在VRF內）到PE2（在VRF內）的eBGP多跳會話，或從PE1（在VRF內）到CE2的eBGP多跳會話，只支援從Cisco IOS®-XR 6.3.2開始。
可通過包含vnpv4的底層到達eBGP對等體地址。vpnv6、6PE或EVPN。
在6.3.2之前的Cisco IOS®版本中，eBGP會話將處於空閒狀態。
例如，在VRF one中配置了eBGP多跳會話PE1到PE2。
從PE1到PE2的eBGP多跳會話的相關配置：
interface Loopback100
vrf one
ipv4 address 10.2.100.1 255.255.255.255
router bgp 1
address-family vpnv4 unicast
!
neighbor 10.100.1.2
remote-as 1
update-source Loopback0
address-family vpnv4 unicast
!
!
vrf one
rd 1:1
address-family ipv4 unicast
redistribute connected
!
neighbor 10.2.100.2
remote-as 65002
ebgp-multihop 255
local-as 65001
update-source Loopback100
address-family ipv4 unicast
route-policy pass in
route-policy pass out
!
!
!
!
eBGP會話保持空閒：
RP/0/0/CPU0:PE1#show bgp vrf one neighbors
BGP neighbor is 10.2.100.2, vrf one
Remote AS 65002, local AS 65001, external link
Remote router ID 0.0.0.0
BGP state = Idle (No route to multi-hop neighbor)
eBGP對等體地址的路由存在於VRF one路由表中：
RP/0/0/CPU0:PE1# show route vrf one
Codes: C - connected, S - static, R - RIP, B - BGP, (>) - Diversion path
D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
E1 - OSPF external type 1, E2 - OSPF external type 2, E - EGP
i - ISIS, L1 - IS-IS level-1, L2 - IS-IS level-2
ia - IS-IS inter area, su - IS-IS summary null, * - candidate default
U - per-user static route, o - ODR, L - local, G - DAGR
A - access/subscriber, a - Application route, (!) - FRR Backup path
Gateway of last resort is not set
L 10.2.100.1/32 is directly connected, 00:23:25, Loopback100
B 10.2.100.2/32 [200/0] via 10.100.1.2 (nexthop in vrf default), 00:19:28
RP/0/0/CPU0:PE1# show route vrf one 10.2.100.2/32
Routing entry for 10.2.100.2/32
Known via "bgp 1", distance 200, metric 0, type internal
Installed May 29 09:07:53.368 for 00:19:36
Routing Descriptor Blocks
10.100.1.2, from 10.100.1.2
Nexthop in Vrf: "default", Table: "default", IPv4 Unicast, Table Id: 0xe0000000
Route metric is 0
No advertising protos.
問題的根本原因是對等地址的路由是匯入的路由：
RP/0/0/CPU0:PE1# show bgp vpnv4 unicast vrf one 10.2.100.2/32
BGP routing table entry for 10.2.100.2/32, Route Distinguisher: 1:1
Versions:
Process bRIB/RIB SendTblVer
Speaker 7 7
Last Modified: May 29 09:07:53.524 for 00:21:20
Paths: (1 available, best #1)
Not advertised to any peer
Path #1: Received by speaker 0
Not advertised to any peer
Local
10.100.1.2 (metric 2) from 10.100.1.2 (10.100.1.2)
Received Label 16001
Origin incomplete, metric 0, localpref 100, valid, internal, best, group-best, import-candidate, imported
Received Path ID 0, Local Path ID 1, version 7
Extended community: RT:1:1
Source VRF: one, Source Route Distinguisher: 1:1
Cisco IOS®-XR 6.3.2後支援此操作。
8.將BGP重新分配到LDP
這就是統一或無縫MPLS的本質及其使用IOS-XR配置的方式：使用IOS-XR的統一MPLS
對於常規的統一MPLS，所有PE路由器和ABR路由器之間都存在BGP LU，如下圖所示。
圖9.
圖10.
在本例中，有一個IGP區域/級別沒有BGP LU。在左側，聚合區域實際上是開放最短路徑優先(OSPF)進程1，該進程在核心沒有使用OSPF進程2進行重分發。在使用OSPF 1的網路部分，PE路由器與區域邊界路由器(ABR)路由器之間沒有BGP LU。
圖11.
BGP LU字首將重分佈到ABR1上的IGP OSPF 1中，如下圖所示。
圖12.
您需要BGP為接收的iBGP LU字首分配標籤。但是，LDP不會在標籤繫結中自動通告此標籤的重新分發字首。預設情況下，IOS(-XE)會執行此操作。
請注意，ABR正在將內部BGP路由重分佈到左側區域的IGP中。這表示路由器bgp下需要命令bgp redistribute-internal。
router bgp 1
bgp redistribute-internal
router ospf 1
router-id 10.100.1.3
redistribute bgp 1 metric 10 route-policy select-to-allocate
area 0
interface Loopback0
!
interface GigabitEthernet0/0/0/0
network point-to-point
!
!
!
route-policy select-to-allocate
if destination in (10.100.1.7/32) then
pass
else
drop
endif
end-policy
啟用本地標籤分配時，ABR會將本地標籤分配給接收到的iBGP LU路由。
router bgp 1
bgp redistribute-internal
ibgp policy out enforce-modifications
address-family ipv4 unicast
redistribute ospf 1 metric 10 route-policy ospf-1-loopbacks-PE
allocate-label route-policy select-to-allocate
route-policy select-to-allocate可用於指定將哪些接收的BGP LU字首分配給本地標籤。
route-policy select-to-allocate
if destination in (10.100.1.7/32) then
pass
else
drop
endif
end-policy
!
ABR1上使用本地標籤顯示PE2的環回字首，但LDP看不到此本地標籤：
RP/0/0/CPU0:ABR1#show bgp ipv4 labeled-unicast 10.100.1.7/32
BGP routing table entry for 10.100.1.7/32
Versions:
Process bRIB/RIB SendTblVer
Speaker 6 6
Local Label: 24006
Last Modified: Sep 5 06:55:47.368 for 06:40:23
Paths: (1 available, best #1)
Advertised IPv4 Labeled-unicast paths to update-groups (with more than one peer):
0.2
Path #1: Received by speaker 0
Advertised IPv4 Labeled-unicast paths to update-groups (with more than one peer):
0.2
Local, (Received from a RR-client)
10.100.1.5 (metric 20) from 10.100.1.5 (10.100.1.7)
Received Label 24003
Origin IGP, metric 0, localpref 100, valid, internal, best, group-best, labeled-unicast
Received Path ID 0, Local Path ID 1, version 6
Originator: 10.100.1.7, Cluster list: 10.100.1.5
RP/0/0/CPU0:ABR1#show mpls ldp bindings 10.100.1.7/32
10.100.1.7/32, rev 0 (no route)
No local binding
Remote bindings: (1 peers)
Peer Label
----------------- ---------
10.100.1.2:0 18
這意味著從PE1到PE2的LSP中斷：
RP/0/0/CPU0:PE1#traceroute 10.100.1.7 source 10.100.1.1
Type escape sequence to abort.
Tracing the route to 10.100.1.7
1 10.1.1.2 [MPLS: Label 18 Exp 0] 9 msec 0 msec 0 msec
2 10.1.2.3 0 msec 0 msec 0 msec <<< no MPLS labels
3 10.1.3.4 [MPLS: Labels 16/24003 Exp 0] 29 msec 19 msec 29 msec
4 10.1.4.5 [MPLS: Label 24003 Exp 0] 9 msec 9 msec 9 msec
5 * * *
6 10.1.6.7 9 msec * 19 msec
LSP在P2被中斷，因為它沒有從ABR1通過LDP獲取遠端標籤。ABR1沒有在LDP中為字首10.100.1.7/32本地分配的標籤。
在BGP路由被重分發到IGP的路由器上，ABR上需要配置將BGP重分發到LDP。
ABR1不會向P2路由器通告字首10.100.1.7/32的LDP標籤繫結。
為了使ABR1通告重新分發的iBGP字首的LDP標籤繫結，ABR1必須具有以下配置（必須配置AS編號）。
mpls ldp
mldp
address-family ipv4
!
!
router-id 10.100.1.3
address-family ipv4
redistribute
bgp
as 1
!
!
!
可以讓LDP過濾通告。例如，您可以按如下方式配置過濾器：
mpls ldp
mldp
address-family ipv4
!
!
router-id 10.100.1.3
address-family ipv4
redistribute
bgp
as 1
advertise-to 1
!
ipv4 access-list 1
10 permit ipv4 host 10.100.1.2 any
您可以在訪問清單中指定LDP router-ID。
在本示例中，ABR只向LDP鄰居P2（而不是向P1）通告重分佈的iBGP路由的LDP繫結，因為10.100.1.2是P2的LDP路由器ID。
從PE1到PE2的LSP現在不受干擾：
RP/0/0/CPU0:PE1#traceroute 10.100.1.7 source 10.100.1.1
Type escape sequence to abort.
Tracing the route to 10.100.1.7
1 10.1.1.2 [MPLS: Label 20 Exp 0] 39 msec 49 msec 29 msec
2 10.1.2.3 [MPLS: Label 24006 Exp 0] 29 msec 49 msec 39 msec
3 10.1.3.4 [MPLS: Labels 16/24003 Exp 0] 29 msec 19 msec 29 msec
4 10.1.4.5 [MPLS: Label 24003 Exp 0] 29 msec 19 msec 29 msec
5 * * *
6 10.1.6.7 19 msec * 19 msec
圖13.
現在，由LDP通告到左側聚合區域的BGP分配標籤(24006)用於從PE1到PE2的流量。
請注意，左側聚合區域中僅使用一個MPLS標籤。如果這是常規統一MPLS，則使用兩個標籤。
此時，您無法過濾哪些重分發的LU iBGP路由進入LDP、接收本地標籤以及哪些沒有。啟用iBGP LU路由到LDP的再分配後，它們都會獲得本地標籤。
PE2還在BGP LU中通告字首10.100.1.99/32。ABR1不會將此字首重新分發到OSPF 1。但是，一旦開啟將iBGP LU路由重新分發到LDP，字首10.100.1.99/32也會獲得本地標籤。
RP/0/0/CPU0:ABR1#show mpls ldp bindings 10.100.1.99/32
10.100.1.99/32, rev 24
Local binding: label: 24007
No remote bindings
9. MPLS啟用介面命令
示例1. IGP，但沒有LDP
如果有負責內部路由的IGP，但沒有用於通告標籤繫結的LDP，則需要使用mpls activate命令。如果每個躍點執行BGP，則BGP LU可用於通告字首和標籤。當它在鏈路上為iBGP時，需要使用mpls activate命令在路由器BGP下啟用該鏈路。請參閱此映像。
圖14.
R1和R2運行它們之間的IGP和iBGP LU。R1和R2直接連線。R2具有到R3的eBGP LU會話。
R3通過eBGP LU會話將字首10.100.100.3/2通告給R2。R2通過iBGP LU會話向R1通告此字首。
目標是有一個從R1到R3的不間斷LSP。它在嗎？
RP/0/0/CPU0:R1#trace 10.100.100.3 so 10.100.100.1
Type escape sequence to abort.
Tracing the route to 10.100.100.3
1 100.1.1 !N * !N
第一跳處沒有此字首的標籤。
RP/0/0/CPU0:R1#traceroute mpls ipv4 10.100.100.3/32 ttl 5
Tracing MPLS Label Switched Path to 10.100.100.3/32, timeout is 2 seconds
Codes: '!' - success, 'Q' - request not sent, '.' - timeout,
'L' - labeled output interface, 'B' - unlabeled output interface,
'D' - DS Map mismatch, 'F' - no FEC mapping, 'f' - FEC mismatch,
'M' - malformed request, 'm' - unsupported tlvs, 'N' - no rx label,
'P' - no rx intf label prot, 'p' - premature termination of LSP,
'R' - transit router, 'I' - unknown upstream index,
'X' - unknown return code, 'x' - return code 0
Type escape sequence to abort.
0 0.0.0.0 MRU 0 [No Label]
Q 1 *
所以，沒有標籤。這並不令人意外，因為到R2的介面上未啟用MPLS:
RP/0/0/CPU0:R1#show mpls interfaces
RP/0/0/CPU0:R1#
但是，R3通告的LU字首存在於R1上：
RP/0/0/CPU0:R1#show bgp ipv4 labeled-unicast 10.100.100.3/32
BGP routing table entry for 10.100.100.3/32
Versions:
Process bRIB/RIB SendTblVer
Speaker 7 7
Local Label: 24001
Last Modified: Sep 13 14:27:17.510 for 00:11:39
Paths: (1 available, best #1)
Not advertised to any peer
Path #1: Received by speaker 0
Not advertised to any peer
65001
10.100.1.2 (metric 2) from 10.100.1.2 (10.100.1.2)
Received Label 24002
Origin IGP, metric 0, localpref 100, valid, internal, best, group-best, labeled-unicast
Received Path ID 0, Local Path ID 1, version 7
您可以在R1上為通向R2的介面配置mpls active命令：
router bgp 65000
mpls activate
interface GigabitEthernet0/0/0/0
!
address-family ipv4 unicast
network 10.100.100.1/32
allocate-label all
!
neighbor 10.100.1.2
remote-as 65000
update-source Loopback0
address-family ipv4 labeled-unicast
!
!
!
現在，在傳出介面上啟用了MPLS。
RP/0/0/CPU0:R1#show mpls interfaces
Interface LDP Tunnel Static Enabled
-------------------------- -------- -------- -------- --------
GigabitEthernet0/0/0/0 No No No Yes
traceroute現在顯示LSP未中斷。
RP/0/0/CPU0:R1#trace 10.100.100.3 so 10.100.100.1
Type escape sequence to abort.
Tracing the route to 10.100.100.3
1 10.1.2.2 [MPLS: Label 24002 Exp 0] 39 msec 9 msec 9 msec
2 10.2.3.3 19 msec * 9 msec
RP/0/0/CPU0:R1#traceroute mpls ipv4 10.100.100.3/32 ttl 5 source 10.100.100.1
Tracing MPLS Label Switched Path to 10.100.100.3/32, timeout is 2 seconds
Codes: '!' - success, 'Q' - request not sent, '.' - timeout,
'L' - labeled output interface, 'B' - unlabeled output interface,
'D' - DS Map mismatch, 'F' - no FEC mapping, 'f' - FEC mismatch,
'M' - malformed request, 'm' - unsupported tlvs, 'N' - no rx labl,
'P' - no rx intf label prot, 'p' - premature termination of LSP,
'R' - transit router, 'I' - unknown upstream index,
'X' - unknown return code, 'x' - return code 0
Type escape sequence to abort.
0 10.1.2.1 MRU 1500 [Labels: implicit-null/24002 Exp: 0/0]
L 1 10.1.2.2 MRU 1500 [Labels: implicit-null/implicit-null Exp: 0/0] 0 ms
! 2 10.2.3.3 10 ms
範例2.聯盟
此範例說明使用BGP LU(RFC 3107)而不使用LDP時，eBGP（AS間）聯盟連結上需要mpls activate命令。
此映像中的網路是包含子65000自治系統65501、65502、65503和介面的聯65504。
圖15.
其思想是通過在兩個自治系統中使用BGP LU，將一個MPLS LSP從R1到R8(10.0.0.8/32由R8在BGP LU中通告)。
R7和R8之間有常規的eBGP LU，R2和R4之間以及R5和R6之間有iBGP，R1和R2、R4和R5之間以及R6和R7之間有eBGP。每個eBGP會話上都有下一跳自身。
需要到達eBGP對等體的下一躍點的靜態路由（通常用於AS BGP間會話），因為聯盟內的子自治系統之間存在eBGP。
這是否足以實現R1和R8之間的連線？這意味著目標是有一個從R1到R8的不間斷LSP。
看看這個。
RP/0/0/CPU0:R1#traceroute 10.0.0.8
Type escape sequence to abort.
Tracing the route to 10.0.0.8
1 * * *
2 * * *
3 * * *
4 * * *
5 * * *
traceroute不會返回任何躍點/標籤，且如果命令上未提供TTL限制，則會繼續。路由器可能應答traceroute，但資料包可能不會返回R1。執行mpls traceroute是一個更安全的選擇。
RP/0/0/CPU0:R1#trace mpls ipv4 10.0.0.8/32
Tracing MPLS Label Switched Path to 10.0.0.8/32, timeout is 2 seconds
Codes: '!' - success, 'Q' - request not sent, '.' - timeout,
'L' - labeled output interface, 'B' - unlabeled output interface,
'D' - DS Map mismatch, 'F' - no FEC mapping, 'f' - FEC mismatch,
'M' - malformed request, 'm' - unsupported tlvs, 'N' - no rx label,
'P' - no rx intf label prot, 'p' - premature termination of LSP,
'R' - transit router, 'I' - unknown upstream index,
'X' - unknown return code, 'x' - return code 0
Type escape sequence to abort.
0 10.1.2.1 MRU 1500 [Labels: implicit-null/24015 Exp: 0/0]
L 1 10.1.2.2 MRU 1500 [Labels: 24003/24014 Exp: 0/0] 10 ms
L 2 10.2.3.3 MRU 1500 [Labels: implicit-null/24014 Exp: 0/0] 10 ms
N 3 10.3.4.4 MRU 0 [No Label] 10 ms
您發現問題出在R4上。LFIB中缺少傳出介面：
RP/0/0/CPU0:R4#show mpls forwarding prefix 10.0.0.8/32
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
24014 24014 10.0.0.8/32 10.4.5.5 5140
未解析CEF中的條目：
RP/0/0/CPU0:R4#show cef 10.0.0.8/32
10.0.0.8/32, version 109, drop adjacency, internal 0x5000001 0x0 (ptr 0xa14160e4) [1], 0x0 (0xa13f83c8), 0xa08 (0xa16cd370)
Updated Sep 13 12:43:30.252
Prefix Len 32, traffic index 0, precedence n/a, priority 4
via 10.4.5.5/32, 0 dependencies, recursive [flags 0x6000]
path-idx 0 NHID 0x0 [0xa0f182d8 0x0]
recursion-via-/32
unresolved
local label 24014
labels imposed {24014}
GE0/0/0/1介面上未啟用MPLS:
RP/0/0/CPU0:R4#show mpls interfaces
Interface LDP Tunnel Static Enabled
-------------------------- -------- -------- -------- --------
GigabitEthernet0/0/0/0 Yes No No Yes
此問題可通過在R4和R5之間的鏈路上啟用BGP的MPLS命令得到解決。R4和R5在此鏈路上具有eBGP聯盟會話。65000實際上，這是聯盟域內的iBGP作業階段。因此，需要使用啟用MPLS的命令來確保R4上的首碼解析到下一個躍點R5。在其他常規網路中，會有LDP處理此作業，但在這裡，R4和R5之間沒有LDP，因為它是在聯盟域內的eBGP作業階段。
為R4上的介面ge 0/0/0/1新增mpls activate命令：
router bgp 65502
bgp confederation peers
65501
65503
65504
!
bgp confederation identifier 65000
mpls activate
interface GigabitEthernet0/0/0/1
!
…
RP/0/0/CPU0:R4#show mpls interfaces
Interface LDP Tunnel Static Enabled
-------------------------- -------- -------- -------- --------
GigabitEthernet0/0/0/0 Yes No No Yes
GigabitEthernet0/0/0/1 No No No Yes
現在traceroute顯示從R1到R8的不連續LSP。
RP/0/0/CPU0:R1#trace mpls ipv4 10.0.0.8/32
Tracing MPLS Label Switched Path to 10.0.0.8/32, timeout is 2 seconds
Codes: '!' - success, 'Q' - request not sent, '.' - timeout,
'L' - labeled output interface, 'B' - unlabeled output interface,
'D' - DS Map mismatch, 'F' - no FEC mapping, 'f' - FEC mismatch,
'M' - malformed request, 'm' - unsupported tlvs, 'N' - no rx label
'P' - no rx intf label prot, 'p' - premature termination of LSP,
'R' - transit router, 'I' - unknown upstream index,
'X' - unknown return code, 'x' - return code 0
Type escape sequence to abort.
0 10.1.2.1 MRU 1500 [Labels: implicit-null/24015 Exp: 0/0]
L 1 10.1.2.2 MRU 1500 [Labels: 24003/24014 Exp: 0/0] 10 ms
L 2 10.2.3.3 MRU 1500 [Labels: implicit-null/24014 Exp: 0/0] 10 ms
L 3 10.3.4.4 MRU 1500 [Labels: implicit-null/24014 Exp: 0/0] 10 ms
L 4 10.4.5.5 MRU 1500 [Labels: implicit-null/24014 Exp: 0/0] 20 ms
L 5 10.5.6.6 MRU 1500 [Labels: implicit-null/24014 Exp: 0/0] 30 ms
L 6 10.6.7.7 MRU 1500 [Labels: implicit-null/implicit-null Exp: 0/0] 30 ms
! 7 10.7.8.8 30 ms
RP/0/0/CPU0:R1#traceroute 10.0.0.8
Type escape sequence to abort.
Tracing the route to 10.0.0.8
1 10.1.2.2 [MPLS: Label 24015 Exp 0] 69 msec 29 msec 29 msec
2 10.2.3.3 [MPLS: Labels 24003/24014 Exp 0] 49 msec 29 msec 29 msec
3 10.3.4.4 [MPLS: Label 24014 Exp 0] 19 msec 19 msec 19 msec
4 10.4.5.5 [MPLS: Label 24014 Exp 0] 49 msec 19 msec 29 msec
5 10.5.6.6 [MPLS: Label 24014 Exp 0] 19 msec 19 msec 29 msec
6 10.6.7.7 [MPLS: Label 24014 Exp 0] 29 msec 29 msec 29 msec
7 10.7.8.8 29 msec * 29 msec
LFIB中現在有一個用於此條目的傳出介面：
RP/0/0/CPU0:R4#show mpls forwarding prefix 10.0.0.8/32
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
24014 24014 10.0.0.8/32 Gi0/0/0/1 10.4.5.5 2890
R4上存在用於字首的傳出標籤，並且CEF顯示解析的字首：
RP/0/0/CPU0:R4#show cef 10.0.0.8/32
Updated Sep 13 12:43:30.252
Prefix Len 32, traffic index 0, precedence n/a, priority 4
via 10.4.5.5/32, 3 dependencies, recursive [flags 0x6000]
path-idx 0 NHID 0x0 [0xa17420e4 0x0]
recursion-via-/32
next hop 10.4.5.5/32 via 24016/0/21
local label 24014
next hop 10.4.5.5/32 Gi0/0/0/1 labels imposed {ImplNull 24014}
意見