[f-nsp] Problem with IPv6 BGP community propagation

[Youssef Bengelloun-Zahr]

Hello community,

I have been struggling with a nice IPv6 BGP problem.

I am using my route reflectors (CER-RT, v5.3.0b) to generate the BGP
prefixes we will be announcing to our upstreams via our edge routers
(MLXe-4, v5.3.0b).

The configuration works just fine for IPv4, but for some reason, my BGP
communities are not propagated when using the same model under IPv6 AFI.

Some outputs :

- I can see that the RR is generating the prefixe, annoucing it to the ER
and taging it with the right BGP community :



SSH at rr01.par01#sh ipv6 bgp routes detail

Total number of BGP Routes: 1

Status A:AGGREGATE B:BEST b:NOT-INSTALLED-BEST C:CONFED_EBGP D:DAMPED

       E:EBGP H:HISTORY I:IBGP L:LOCAL M:MULTIPATH m:NOT-INSTALLED-MULTIPATH

       S:SUPPRESSED F:FILTERED s:STALE

1       Prefix: 2a00:bf40::/32,  Status: BL,  Age: 0h4m29s

         NEXT_HOP: 100::1, Learned from Peer: Local Router

          LOCAL_PREF: 100,  MED: 1,  ORIGIN: igp,  Weight: 32768

         AS_PATH:

            *COMMUNITIES: 57928:106*

            Adj_RIB_out count: 3,  Admin distance 1

            Tag: 0x0000006a



- The ER receives the BGP announcement and installs in its RIB. But for
some reason, BGP community has disappeared along the way :

er01.par01#sh ipv6 bgp routes detail

Total number of BGP Routes: 1

Status A:AGGREGATE B:BEST b:NOT-INSTALLED-BEST C:CONFED_EBGP D:DAMPED

       E:EBGP H:HISTORY I:IBGP L:LOCAL M:MULTIPATH m:NOT-INSTALLED-MULTIPATH

       S:SUPPRESSED F:FILTERED s:STALE

1       Prefix: 2a00:bf40::/32,  Status: BI,  Age: 0h2m11s

         NEXT_HOP: 100::1, Metric: 1,  Learned from Peer: 2a00:bf40::1
(57928)

          LOCAL_PREF: 100,  MED: 1,  ORIGIN: igp,  Weight: 0

         *AS_PATH:*



- I, of course, have activated the propagation of BGP communities under the
RR's BGP configuration (showing only what's relevant) :

router bgp

....

neighbor ER-PEER-V6 peer-group

neighbor ER-PEER-V6 remote-as 57928

neighbor ER-PEER-V6 description EDGE-ROUTERS-PEERS-V6

neighbor ER-PEER-V6 update-source loopback 1

neighbor ER-PEER-V6 soft-reconfiguration inbound

....

neighbor 2a00:bf40::XYZ peer-group ER-PEER-V6

neighbor 2a00:bf40::XYZ peer-group ER-PEER-V6

....



address-family ipv6 unicast

redistribute static route-map FROM-STATIC-V6-TO-BGP

* neighbor RR-PEER-V6 send-community*

neighbor 2a00:bf40::XYZ activate

neighbor 2a00:bf40::XYZ activate

exit-address-family



No filters, ACLs or route-maps of any kind.

Any ideas ?

Best regards.

Y.

-- 
Youssef BENGELLOUN-ZAHR
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/foundry-nsp/attachments/20120905/183d78a8/attachment.html>