Network design, development and automation for enterprise, service provider and
data center fabrics.

basondole.io

BGP Unnumbered

BGP requires a next-hop for a route to be valid. In IPv4, except for getting the MAC address to put on the packet, the next-hop IP address is not used in the packet at all. In case of IPv6, as well, the next-hop IPv6 address is used to identify the next-hop MAC address, using IPv6’s equivalent of ARP: Neighbor Discovery (ND). Even in IPv6, forwarding to the original destination involves only the next-hop’s MAC address. Therefore the next-hop IP address is used only to get the next-hop’s MAC address.

RFC 5549 builds on this observation and provides an encoding scheme to allow a router to advertise IPv4 routes with an IPv6 next-hop. RFC 5549 is a somewhat obscure RFC, invented in the early years of a new century. Its purpose is to allow the advertisement of an IPv4 route and routing of an IPv4 packet over a pure IPv6 network.

To summarize:
  • BGP unnumbered uses the interface’s IPv6 LLA to set up a BGP session with a peer.
  • The IPv6 LLA of the remote end is discovered via IPv6’s Router Advertisement (RA) protocol.
  • RA provides not only the remote end’s LLA, but also its corresponding MAC address.
  • BGP uses RFC 5549 to encode IPv4 routes as reachable over an IPv6 nexthop, using the IPv6 LLA as the nexthop.

Cumulus implementation

The RIB process programs a static ARP entry with a reserved IPv4 LLA, 169.254.0.1, with the MAC address set to the one learned via RA. BGP hands down to the RIB process IPv4 routes with the IPv6 LLA as the next-hop. The RIB process converts the next-hop to 169.254.0.1 and the outgoing interface before programming the route in the forwarding table as shown here 
cumulus@basondole-leaf:~$ arp
Address                  HWtype  HWaddress           Flags Mask            Iface
169.254.0.1              ether   50:00:00:05:00:01   CM                    swp1

cumulus@basondole-leaf:~$ ip route
192.168.1.1 via 169.254.0.1 dev swp1  proto bgp  metric 20 onlink

cumulus@basondole-spine:~$ net sh int swp1
    Name  MAC                Speed  MTU   Mode
--  ----  -----------------  -----  ----  -------------
UP  swp1  50:00:00:05:00:01  1G     1500  NotConfigured


LLDP Details
------------
LocalPort  RemotePort(RemoteHost)
---------  ----------------------
swp1       swp1(basondole-leaf)

Routing
-------
  Interface swp1 is up, line protocol is up
  .
  .
  Type: Ethernet
  HWaddr: 50:00:00:05:00:01
  inet6 fe80::5200:ff:fe05:1/64
  Hosts use stateless autoconfig for addresses.
  Neighbor address(s):
  inet6 fe80::5200:ff:fe04:1/128

Cumulus configuration

+--------------+                                +---------------+
|              |                                |               |
|    spine     |swp1------------------------swp1|    leaf       |
|              |                                |               |
+--------------+                                +---------------+
Note: By default the ports on cumulus are IPv6 enabled and with BGP unnumbered we do not need IPv4 address on the interfaces. Therefore there is nothing to configure on the swp1 interfaces
Spine
net add loopback lo ip address 192.168.1.1/32
net add bgp autonomous-system 64515
net add bgp network 192.168.1.1/32
net add bgp neighbor swp1 interface remote-as external
net add bgp neighbor swp1 capability extended-nexthop
net commit
leaf
net add loopback lo ip address 192.168.2.1/32
net add bgp autonomous-system 64512
net add bgp network 192.168.2.1/32
net add bgp neighbor swp1 interface remote-as external
net add bgp neighbor swp1 capability extended-nexthop
net commit

Verfication on the spine

We are learning 1 prefix from the neighbor basondole-leaf 
cumulus@basondole-spine:~$ net sh bgp summary

show bgp ipv4 unicast summary
=============================
BGP router identifier 192.168.1.1, local AS number 64515 vrf-id 0
BGP table version 2
RIB entries 3, using 456 bytes of memory
Peers 1, using 19 KiB of memory

Neighbor             V         AS MsgRcvd MsgSent   TblVer  InQ OutQ  Up/Down State/PfxRcd
basondole-leaf(swp1) 4      64512     100     100        0    0    0 00:04:48            1

Total number of neighbors 1

cumulus@basondole-spine:~$ net sh bgp

show bgp ipv4 unicast
=====================
BGP table version is 2, local router ID is 192.168.1.1
Status codes: s suppressed, d damped, h history, * valid, > best, = multipath,
              i internal, r RIB-failure, S Stale, R Removed
Origin codes: i - IGP, e - EGP, ? - incomplete

   Network          Next Hop            Metric LocPrf Weight Path
*> 192.168.1.1/32   0.0.0.0                  0         32768 i
*> 192.168.2.1/32   swp1                     0             0 64512 i

Displayed  2 routes and 2 total paths

cumulus@basondole-spine:~$ net sh route   

show ip route
=============
Codes: K - kernel route, C - connected, S - static, R - RIP,
       O - OSPF, I - IS-IS, B - BGP, E - EIGRP, N - NHRP,
       T - Table, v - VNC, V - VNC-Direct, A - Babel, D - SHARP,
       F - PBR,
       > - selected route, * - FIB route

C>* 192.168.1.1/32 is directly connected, lo, 00:06:49
B>* 192.168.2.1/32 [20/0] via fe80::5200:ff:fe04:1, swp1, 00:06:38



cumulus@basondole-spine:~$ ping 192.168.2.1 -s 192.168.1.1 -c 5
PING 192.168.2.1 (192.168.2.1) 192(220) bytes of data.
200 bytes from 192.168.2.1: icmp_seq=1 ttl=64 time=1.18 ms
200 bytes from 192.168.2.1: icmp_seq=2 ttl=64 time=0.805 ms
200 bytes from 192.168.2.1: icmp_seq=3 ttl=64 time=1.05 ms
200 bytes from 192.168.2.1: icmp_seq=4 ttl=64 time=1.04 ms
200 bytes from 192.168.2.1: icmp_seq=5 ttl=64 time=1.06 ms

--- 192.168.2.1 ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 4013ms
rtt min/avg/max/mdev = 0.805/1.030/1.180/0.124 ms
cumulus@basondole-spine:~$

Verfication on the leaf

cumulus@basondole-leaf:~$ net sh bgp summ

show bgp ipv4 unicast summary
=============================
BGP router identifier 192.168.2.1, local AS number 64512 vrf-id 0
BGP table version 2
RIB entries 3, using 456 bytes of memory
Peers 1, using 19 KiB of memory

Neighbor              V         AS MsgRcvd MsgSent   TblVer  InQ OutQ  Up/Down State/PfxRcd
basondole-spine(swp1) 4      64515      91      92        0    0    0 00:04:20            1

Total number of neighbors 1

cumulus@basondole-leaf:~$ net sh bgp 

show bgp ipv4 unicast
=====================
BGP table version is 2, local router ID is 192.168.2.1
Status codes: s suppressed, d damped, h history, * valid, > best, = multipath,
              i internal, r RIB-failure, S Stale, R Removed
Origin codes: i - IGP, e - EGP, ? - incomplete

   Network          Next Hop            Metric LocPrf Weight Path
*> 192.168.1.1/32   swp1                     0             0 64515 i
*> 192.168.2.1/32   0.0.0.0                  0         32768 i

Displayed  2 routes and 2 total paths

Interoperability between vendors

+--------------+                                +---------------+
|              |                                |               |
|    cisco     | Gi2-----------------------swp2 |    cumulus    |
|              |                                |               |
+--------------+                                +---------------+
With cisco IOS XE there’s more to be configured. In BGP configuration you must specify the link local address of the neighbor together with a route-map to change the next-hop address from ipv6 linklocal to ipv4 address. 
Cat8000v(config)#do sh ipv6 int g2 | i IPv6
  IPv6 is enabled, link-local address is FE80::5200:FF:FE06:1 

Cat8000v(config)#do sh ipv6 nei
IPv6 Address                              Age Link-layer Addr State Interface
FE80::5200:FF:FE03:2                        0 5000.0003.0002  REACH Gi2


cumulus@basondole-cu:~$ net sh int swp2 | egrep "Int.*swp|inet|Nei" 
IP Neighbor(ARP) Entries:  2
  Interface swp2 is up, line protocol is up
  inet 10.10.20.2/30
  inet6 fe80::5200:ff:fe03:2/64
  Neighbor address(s):
  inet6 fe80::5200:ff:fe06:1/128

Configuration on IOS XE

interface GigabitEthernet2
 description toCumulus
 ip address 10.10.20.1 255.255.255.252
 ipv6 enable

interface Loopback0
 ip address 192.168.30.1 255.255.255.255

router bgp 64520
 bgp log-neighbor-changes
 neighbor FE80::5200:FF:FE03:2%GigabitEthernet2 remote-as 64512
 !
 address-family ipv4
  network 192.168.30.1 mask 255.255.255.255
  neighbor FE80::5200:FF:FE03:2%GigabitEthernet2 activate
  neighbor FE80::5200:FF:FE03:2%GigabitEthernet2 route-map FRR out
 exit-address-family

route-map FRR permit 10 
 set ip next-hop 10.10.20.1

Configuration on cumulus VX

net add interface swp2 ip address 10.100.10.2/30
net add loopback lo ip address 192.168.20.1/32
net add bgp autonomous-system 64512
net add bgp network 192.168.20.1/32
net add bgp neighbor swp2 interface remote-as external
net add bgp neighbor swp2 capability extended-nexthop
net add routing route-map FRR permit 10 set ip next-hop 10.10.20.2
net add bgp neighbor swp2 route-map FRR out
net commit
In case you observe log messages on IOS XE 
*Dec 29 14:24:02.791: %BGP-5-NBR_RESET: Neighbor 10.10.20.2 passive reset (BGP Notification sent)
*Dec 29 14:24:02.792: %BGP-5-ADJCHANGE: neighbor 10.10.20.2 passive Down Error during connection collision
Remove ipv4 address on swp2, bgp session will come up then restore the ipv4 address, as it is needed for next hop resolution by the cumulus switch

Verification

Cat8000v(config)#do sh bgp summ | b N
Neighbor        V           AS MsgRcvd MsgSent   TblVer  InQ OutQ Up/Down  State/PfxRcd
FE80::5200:FF:FE03:2%GigabitEthernet2
                4        64512     141     139        3    0    0 00:06:53        1


Cat8000v(config)#do sh bgp | b Net
     Network          Next Hop            Metric LocPrf Weight Path
 *>   192.168.20.1/32  10.10.20.2               0             0 64512 i
 *>   192.168.30.1/32  0.0.0.0                  0         32768 i

Cat8000v(config)#do ping 192.168.20.1 so 192.168.30.1
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 192.168.20.1, timeout is 2 seconds:
Packet sent with a source address of 192.168.30.1 
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 1/7/18 ms

Cat8000v(config)#do traceroute 192.168.20.1 so 192.168.30.1 numeric
Type escape sequence to abort.
Tracing the route to 192.168.20.1
VRF info: (vrf in name/id, vrf out name/id)
  1 192.168.20.1 [AS 64512] 1 msec 1 msec 2 msec
Cat8000v(config)#

cumulus@basondole-cu:~$ net sh bgp summary

show bgp ipv4 unicast summary
=============================
BGP router identifier 192.168.20.1, local AS number 64512 vrf-id 0
BGP table version 6
RIB entries 3, using 456 bytes of memory
Peers 4, using 77 KiB of memory

Neighbor        V         AS MsgRcvd MsgSent   TblVer  InQ OutQ  Up/Down State/PfxRcd
swp2            4      64520     177     180        0    0    0 00:08:49            1

Total number of neighbors 1


cumulus@basondole-cu:~$ net sh bgp 

show bgp ipv4 unicast
=====================
BGP table version is 6, local router ID is 192.168.20.1
Status codes: s suppressed, d damped, h history, * valid, > best, = multipath,
              i internal, r RIB-failure, S Stale, R Removed
Origin codes: i - IGP, e - EGP, ? - incomplete

   Network          Next Hop            Metric LocPrf Weight Path
*> 192.168.20.1/32  0.0.0.0                  0         32768 i
*> 192.168.30.1/32  10.10.20.1               0             0 64520 i

Displayed  2 routes and 2 total paths


cumulus@basondole-cu:~$ ping 192.168.30.1 -s 192.168.20.1  -c 2
PING 192.168.30.1 (192.168.30.1) 192(220) bytes of data.
200 bytes from 192.168.30.1: icmp_seq=1 ttl=255 time=0.733 ms
200 bytes from 192.168.30.1: icmp_seq=2 ttl=255 time=0.775 ms

--- 192.168.30.1 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1001ms
rtt min/avg/max/mdev = 0.733/0.754/0.775/0.021 ms


cumulus@basondole-cu:~$ traceroute 192.168.30.1 -s 192.168.20.1 -n
traceroute to 192.168.30.1 (192.168.30.1), 30 hops max, 60 byte packets
 1  10.10.20.1  1.638 ms * *
cumulus@basondole-cu:~$
Juniper configuration
basondole@vQFX# show | compare
[edit groups]
+  BGP_UNNUMBERED {
+      interfaces {
+          xe-0/0/1 {
+              unit 0 {
+                  description to_GNS3_via_vboxNet3;
+                  family inet {
+                      address 172.17.0.1/30;
+                  }
+                  family inet6;
+              }
+          }
+      }
+      protocols {
+          bgp {
+              group FRR {
+                  family inet {
+                      unicast;
+                  }
+                  peer-as 64515;
+                  local-as 64514;
+                  neighbor fe80::5200:ff:fe05:1 {
+                      local-interface xe-0/0/1.0;
+                  }
+              }
+          }
+      }
+  }
[edit]
+ apply-groups BGP_UNNUMBERED;

Paul S.I. Basondole

IT & Network Expert / Lead
Network & Automation designer;
2x JNCIP (RS/SP) 1x CCNP (RS)
BSc Telecommunications Eng.
© 2019 basondole.io

Sliding Sidebar