[c-nsp] "continue" clause in route-maps
Pete Templin
petelists at templin.org
Tue Feb 15 09:45:11 EST 2011
On 2/14/2011 1:03 PM, Yann GAUTERON wrote:
> Hi !
>
> I would like to apply some sophisticated rules with my prefixes announced
> with BGP.
>
> I saw that Cisco implemented a "continue" clause that could permit me to
> achieve my goal.
>
> Before implemented such rules on my productive routers, I did some tests
> with GNS3, but it seems that the "continue" does not work !
>
> For a simple simulation with 2 routers (in BGP AS 1 and AS 2), I cannot see
> the expected result.
>
> Configuration on the router in AS1:
> ip cef
> !
> interface Loopback0
> ip address 1.1.1.1 255.255.255.255
> !
> interface GigabitEthernet1/0
> ip address 10.0.0.1 255.255.255.0
> negotiation auto
> !
> router bgp 1
> no synchronization
> bgp log-neighbor-changes
> network 1.0.0.0 route-map NET-1-0-0-IN
> neighbor 10.0.0.2 remote-as 2
> neighbor 10.0.0.2 send-community
> neighbor 10.0.0.2 route-map PEER-OUT out
> no auto-summary
> !
> ip forward-protocol nd
> ip route 1.0.0.0 255.0.0.0 Null0
> !
> ip bgp-community new-format
> ip community-list 1 permit 1:1
> !
> route-map PEER-OUT permit 10
> match community 1
> continue
> set community 1:100 additive
> !
> route-map PEER-OUT permit 20
> set community 1:200 additive
> !
> route-map NET-1-0-0-IN permit 10
> set community 1:1 1:2 65000:1 65000:2
>
> With this configuration, I would expect my second router (in AS 2), to
> receive the prefix 1.0.0.0/8 with the communities 1:1 1:2 1:100 1:200
> 65000:1 65000:2.
>
> But I cannot see 1:200 in my communities:
> Router#sh ip bgp 1.0.0.0
> BGP routing table entry for 1.0.0.0/8, version 9
> Paths: (1 available, best #1, table Default-IP-Routing-Table)
> Flag: 0x820
> Not advertised to any peer
> 1
> 10.0.0.1 from 10.0.0.1 (1.1.1.1)
> Origin IGP, metric 0, localpref 100, valid, external, best
> Community: 1:1 1:2 1:100 65000:1 65000:2
>
> My GNS3 router is a 7200 and has IOS 12.4(25d). My productive router is a
> 7600 with 12.2(33)SRE, but I cannot emulate a 7600 on GNS3 !
>
> Is the problem due to a misbehaviour of that IOS 12.4(25d) on my "virtual"
> 7200, or did I miss something with the way "continue" do work ? Is it a
> misbehavior I would face with my productive 7600 ?
>
For the sake of asking, are you certain that BRMCO is supported in
12.4(25d) on the 7200 platform? Note that outbound support followed
long after inbound support, and outbound was unstable when first
released in 12.0(31)S (I filed a bug on it, and it wasn't fixed until
12.0(32)S). Try putting a 'set metric 12345' in the PEER-OUT p 20
clause and see if that takes effect. If it does, you have a
misbehavior. If it doesn't, BRMCO doesn't work on this code.
For the sake of sane troubleshooting, your configuration snippet above
is misleading, and overly complex. If you're going to simulate BRMCO,
just simulate BRMCO - add communities on the outbound side and see what
you get on the inbound side. In your example above, you have a
two-clause outbound route map where each clause adds one community. You
then have an "unused" inbound route-map which sets four communities
(non-additive, I might add). If you're trying to test the outbound
behavior, just test that. If you're trying to show the inbound
behavior, we're going to tell you to isolate the one thing you're trying
to do so we can see it. Also, if you're trying to show us that
route-map NET-1-0-0-IN is what you're applying on the inbound router,
show us a separate config for that router with that route-map. If
you're trying to suggest that this route-map is being used inbound, it's
not clear above. If I were to guess that's what you were doing, I'd
tell you you'll never see communities 1:100 or 1:200 because your
route-map inbound isn't additive on the communities.
pt
More information about the cisco-nsp
mailing list