[c-nsp] ipv6 tunnels and static routes

Pierfrancesco Caci p.caci at seabone.net
Thu Mar 31 08:49:48 EST 2005



Hello,
I've been banging my head all day with this problem. 

Background: a 6PE network, static and connected routes are
redistributed into iBGP with a route-map which applies no-export and
does some sanity filtering.

I have a tunnel from a GSR towards a linux machine, and I've assigned
a /48 to the "customer" using a static route. These are the relevant
configurations  


interface Tunnel1001
 description tunnel IPv6 Test Pf
 no ip address
 no ip directed-broadcast
 load-interval 30
 ipv6 address 2001:41A8:600:2::5/126
 tunnel source Loopback0
 tunnel destination x.x.141.114
 tunnel mode ipv6ip
end

On the linux machine we have:

auto seabone-decix
iface seabone-decix inet6 v4tunnel
        address 2001:41A8:600:2::6
        netmask 126
        local x.x.141.114
        endpoint 195.22.211.248
        ttl 128

which yields:

seabone-decix: ipv6/ip  remote 195.22.211.248  local x.x.141.114  ttl 128

51: seabone-decix at NONE: <POINTOPOINT,NOARP,UP> mtu 1480 qdisc noqueue 
    link/sit x.x.141.114 peer 195.22.211.248
    inet6 2001:41a8:600:2::6/126 scope global 
    inet6 fe80::c0a8:206/128 scope link 

default via 2001:41a8:600:2::5 dev seabone-decix  proto zebra  metric 1024  mtu 1480 advmss 1420


On the 12000 we get:

G12406A-FR_INT01#sh ipv6 route 2001:41a8:604:: 
IPv6 Routing Table - 334 entries
Codes: C - Connected, L - Local, S - Static, R - RIP, B - BGP
       I1 - ISIS L1, I2 - ISIS L2, IA - ISIS interarea
       O - OSPF intra, OI - OSPF inter, OE1 - OSPF ext 1, OE2 - OSPF ext 2
S   2001:41A8:604::/48 [1/0]
     via 2001:41A8:600:2::6, Tunnel1001, 00:16:12


G12406A-FR_INT01#sh bgp ipv6 2001:41A8:604::/48
BGP routing table entry for 2001:41A8:604::/48, version 159784
Paths: (1 available, best #1, table Global-IPv6-Table, not advertised to EBGP peer)
  Advertised to update-groups:
     1         
  Local
    2001:41A8:600:2::6 from 0.0.0.0 (195.22.211.248)
      Origin incomplete, metric 0, localpref 100, weight 32768, valid, sourced, best
      Community: no-export


On another router of the net we can see:

R_7206A-MI_MM_19#sh ipv6 route  2001:41A8:604::1
IPv6 Routing Table - 336 entries
Codes: C - Connected, L - Local, S - Static, R - RIP, B - BGP
       U - Per-user Static route
       I1 - ISIS L1, I2 - ISIS L2, IA - ISIS interarea, IS - ISIS summary
       O - OSPF intra, OI - OSPF inter, OE1 - OSPF ext 1, OE2 - OSPF ext 2
       ON1 - OSPF NSSA ext 1, ON2 - OSPF NSSA ext 2
B   2001:41A8:604::/48 [200/0]
     via ::FFFF:195.22.211.248, IPv6-mpls


R_7206A-MI_MM_19#sh bgp ipv6 2001:41A8:604::/48  
BGP routing table entry for 2001:41A8:604::/48, version 327808
Paths: (2 available, best #2, table Global-IPv6-Table, not advertised to EBGP peer)
  Not advertised to any peer
  Local
    ::FFFF:195.22.211.248 (metric 121) from 195.22.208.252 (195.22.208.252)
      Origin incomplete, metric 0, localpref 100, valid, internal
      Community: 6762:92 no-export
      Originator: 195.22.211.248, Cluster list: 0.0.0.1, 0.0.0.4
  Local
    ::FFFF:195.22.211.248 (metric 121) from 195.22.208.225 (195.22.208.225)
      Origin incomplete, metric 0, localpref 100, valid, internal, best
      Community: 6762:92 no-export
      Originator: 195.22.211.248, Cluster list: 0.0.0.1, 0.0.0.4

Pings and traceroutes done on the GSR where the tunnel is defined work properly:

G12406A-FR_INT01#ping 2001:41A8:604::1         

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 2001:41A8:604::1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 260/289/316 ms

G12406A-FR_INT01#trace 2001:41A8:604::1

Type escape sequence to abort.
Tracing the route to 2001:41A8:604::1

  1 2001:41A8:600:2::6 268 msec 280 msec 280 msec
  2 2001:41A8:604::1 280 msec 320 msec 248 msec


...while from another router:


R_7206A-MI_MM_19#ping 2001:41A8:604::1  

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 2001:41A8:604::1, timeout is 2 seconds:
.....
Success rate is 0 percent (0/5)

R_7206A-MI_MM_19#trace 2001:41A8:604::1  

Type escape sequence to abort.
Tracing the route to 2001:41A8:604::1

  1 2001:41A8:600::1E 12 msec 8 msec 12 msec
  2  *  *  * 


The last responding hop is the loopback of the GSR.



Tag-switching infos on the 12000:

G12406A-FR_INT01#sh ipv6 cef tunnel 1001
2001:41A8:600:2::4/126
     attached to Tunnel1001 
  tag information 
    local tag: aggregate(902)
2001:41A8:604::/48
     nexthop 2001:41A8:600:2::6 Tunnel1001 
  tag information 
    local tag: dynamic(915)

G12406A-FR_INT01#sh bgp ipv6 tags 
   Network          Next Hop      In tag/Out tag
[....]
   2001:41A8:600:2::4/126
                    ::              902/notag
   2001:41A8:604::/48
                    2001:41A8:600:2::6
                                    915/notag
[....]


on the other router:

R_7206A-MI_MM_19#sh bgp ipv6 tag
   Network          Next Hop      In tag/Out tag
[....]
   2001:41A8:600:2::4/126
                    ::FFFF:195.22.211.248
                                    notag/902
                    ::FFFF:195.22.211.248
                                    notag/902
   2001:41A8:604::/48
                    ::FFFF:195.22.211.248
                                    notag/915
                    ::FFFF:195.22.211.248
                                    notag/915
[....]


so it seems that tags match properly


R_7206A-MI_MM_19#sh ipv6 cef 2001:41A8:604::/48
2001:41A8:604::/48
     nexthop ::FFFF:195.22.211.248
    fast tag rewrite with Fa0/0, 195.22.208.2, tags imposed: {440 915}



The GSR is running 12.0(26)S5, while the other router has 12.3(7)T6

I found a few bugs (looking for ipv6 and 12.0(26)) that could match,
but they seem to be resolved for 12.0(26)S5, unless I totally
misunderstood them. 

One of those in particular, CSCdy84998, mentions route reflector
clients. The GSR in this context is a RRC of another router. 

I have another similar setup on the abovementioned 7206 where the
static route is correctly working and is reachable from
outside. Comparing the iBGP informations for the 2 statics doesn't
give any hint, as they look the same. 

Anyone that can help ?

Thanks

Pf

-- 


-------------------------------------------------------------------------------
 Pierfrancesco Caci | Network & System Administrator - INOC-DBA: 6762*PFC
 p.caci at seabone.net | Telecom Italia Sparkle - http://etabeta.noc.seabone.net/
     Linux paperino 2.4.20 #1 Mon Dec 2 17:02:14 CET 2002 i686 GNU/Linux



More information about the cisco-nsp mailing list