IPv6 Routing & Redistribution Problem - Not Learning All Subnets on one side

[Dean Romanelli]

Hi All,

I am building my own interpretation of the IPv6 topology used in the CCNP TSHOOT exam in my basement for practice purposes.  This topology is made public by Cisco, and the configurations are my interpretation, NOT a copy of what they use, so there is no NDA problem here. 

I am having a problem pinging from my distribution layer 3 switches (DSW1 & DSW2) to my far-end router (R1), and it doesn't make sense to me. Otherwise, everything else works fine.  If I put a static route in the distribution switch for the far-end R1 network, all works fine, but I shouldn't have to do that because I am redistributing IPv6 RIP into IPv6 OSPF and vice-versa on R4 (next hop router from DSW's), so my Distribution switches should learn about the OSPF routes from external IPv6 RIP routes from the redistribution:

Topology is attached.

Here are my DSW IPv6 rip configs:

DSW1#show run
Building configuration...

Current configuration : 8863 bytes
!
interface FastEthernet1/0/1
 no switchport
 ipv6 address 2026::2:2/122
 ipv6 enable
 ipv6 rip RIP_ZONE enable
 no mdix auto
!
interface FastEthernet1/0/11
 no switchport
 no ip address
 ipv6 address 2026::3:1/122
 ipv6 enable
 ipv6 rip RIP_ZONE enable
 no mdix auto
!
ipv6 router rip RIP_ZONE
!
ipv6 unicast-routing

DSW2#show run
Building configuration...

Current configuration : 8269 bytes
!
interface FastEthernet0/11
 no switchport
 no ip address
 ipv6 address 2026::3:2/122
 ipv6 enable
 ipv6 rip RIP_ZONE enable
!
ipv6 router rip RIP_ZONE

Here are my Redistribution Configs on R4 between ipv6 rip and ospf:

R4#show run
Building configuration...

Current configuration : 4728 bytes
!
ipv6 router ospf 6
 router-id 4.4.4.4
 log-adjacency-changes
 redistribute rip RIP_ZONE include-connected
!
ipv6 router rip RIP_ZONE
 redistribute ospf 6 include-connected

Far-End R1 Router IPv6 Routing Table showing it is learning the DSW routes from said Redistribution:

R1#show ipv6 route
IPv6 Routing Table - 8 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
OI  2026::1:0/122 [110/130]
     via FE80::12:2, Serial0/0/0.12
OE2  2026::2:0/122 [110/20]
     via FE80::12:2, Serial0/0/0.12
OE2  2026::3:0/122 [110/20]
     via FE80::12:2, Serial0/0/0.12
C   2026::12:0/122 [0/0]
     via ::, Serial0/0/0.12
L   2026::12:1/128 [0/0]
     via ::, Serial0/0/0.12
OI  2026::34:0/122 [110/11241]
     via FE80::12:2, Serial0/0/0.12
L   FE80::/10 [0/0]
     via ::, Null0
L   FF00::/8 [0/0]
     via ::, Null0

But Can't Ping them from R1:

R1#ping ipv6 2026::2:2

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 2026::2:2, timeout is 2 seconds:
.....
Success rate is 0 percent (0/5)

And That is because the IPv6 Routing Table on DSW1 isn't learning the R1 IPv6 route for R1-R2 link (2026::12:/122):

But what is weird is it does learn the 2026::34:/122 GRE tunnel network on R4, but nothing else outside of the DSW's:

DSW1#show ipv6 route
IPv6 Routing Table - 7 entries
Codes: C - Connected, L - Local, S - Static, U - Per-user Static route
       R - RIP
       O - OSPF Intra, OI - OSPF Inter, OE1 - OSPF ext 1, OE2 - OSPF ext 2
       ON1 - OSPF NSSA ext 1, ON2 - OSPF NSSA ext 2
C   2026::2:0/122 [0/0]
     via ::, FastEthernet1/0/1
L   2026::2:2/128 [0/0]
     via ::, FastEthernet1/0/1
C   2026::3:0/122 [0/0]
     via ::, FastEthernet1/0/11
L   2026::3:1/128 [0/0]
     via ::, FastEthernet1/0/11
R   2026::34:0/122 [120/2]
     via FE80::219:E7FF:FE56:7732, FastEthernet1/0/1
L   FE80::/10 [0/0]
     via ::, Null0
L   FF00::/8 [0/0]
     via ::, Null0

Now, as mentioned, if I put a static route into DSW1 for 2026::12:/122 (R1) and point to the next hop IP, everything is fine, but I shouldn't have to do that because I am redistributing on R4.  I could just leave the static route in there and the lab is fully functional, but it's going to annoy me if I don't know why this is happening first.

The only thing I can think of is I am not using a default route on the DSW switches, but again, I shouldn't have to because I am using dynamic routing and including all subnets under the respective dynamic routing instance.

What am I missing here?

[Julio E. Moisa]

Hi

Have you tried to ping the destination from specific source?

example:

ping ipv6 2001::1 source 2001::1:45

[Jon Marshall · Accepted solution]

It learns the GRE tunnel subnet because that is directly connected and you have "include-connected" which you still need but you also need to include a metric otherwise the other OSPF subnets won't be redistributed ie. under your router RIP -

"redistribute ospf 6 include-connected metric <hop count>" 

Jon

[Dean Romanelli]

Thanks Jon, I am sure that is the case.

Question: How do I know what the hop count should be?  I assume 2, since RIPng on R4 exists on DSW1 & DSW2 and DSW2 is 2 hops away from R4?  Though I wonder if it is really 1?  See below:

R4---------DSW1---------DSW2

From R4, 1 hop is on the right side of DSW1 above. A 2nd hop would be to the right side of DSW2, but that side of DSW2 above is not a RIP member. Only the left side of DSW2 that faces DSW1 is a RIP member, and that wouldn't count as a hop, since the right side of DSW1 is the hop for that subnet already.

[Jon Marshall]

You don't necessarily need to worry too much about the exact hop count because all the OSPF subnets are only reachable via R4 anyway so it isn't really going to matter too much.

Just bear in mind though that RIP has a maximum hop count of 15 with 16 being an unreachable network and I'm not aware it is any different with IPv6 so just pick a low number and you should be fine.

Jon

[Dean Romanelli]

Thanks Jon, much appreciated.