[c-nsp] RTBH and MPLS network
Oliver Boehmer (oboehmer)
oboehmer at cisco.com
Fri Nov 25 12:11:44 EST 2005
Daemen, Seth, VF-NL <mailto:Seth.Daemen at vodafone.com> wrote on Friday,
November 25, 2005 3:10 PM:
> What do you mean with: You need to set the next-hop inbound or
> outbound at your vpnv4 neighbor.
> Maybe a stupid question but I'm a student and I have not much
> experience with mpls VPN's.
assuming you have configured "neighbor x.x.x.x send-community both" on
all your neighbors within "address-family vpnv4":
route-map static2bgp permit 10
match tag 666
set local-preference 200
set origin igp
set community 1:1
route-map static2bgp permit 20
...
router bgp XX
address-family ipv4 vrf XXX
redistribute static static2bgp
and then
ip community-list 1 permit 1:1
!
route-map blackhole permit 10
match community 1
set ip next-hop 192.168.0.1
route-map blackhole permit 20
!
router bgp XX
address-family vpnv4
neighbor x.x.x.x route-map blackhole out
this requires that all PE's as well as the vpnv4 route-reflectors (if
present) know about the 192.168.0.1 prefix in the global routing table..
oli
>
>
> Seth
> -----Original Message-----
> From: Oliver Boehmer (oboehmer) [mailto:oboehmer at cisco.com]
> Sent: vrijdag 25 november 2005 14:57
> To: Daemen, Seth, VF-NL; cisco-nsp at puck.nether.net
> Subject: RE: [c-nsp] RTBH and MPLS network
>
> Daemen, Seth, VF-NL <mailto:Seth.Daemen at vodafone.com> wrote on Friday,
> November 25, 2005 2:38 PM:
>
>> No I thinks not that RTBH behaves differently in MPLS networks.
>> Butt I have configuration problems.
>>
>> In the bgp configuration is a redistribute:
>>
>> static route-map black-hole-trigger
>>
>> The following route map is created:
>>
>> route-map black-hole-trigger, permit, sequence 10
>> Match clauses:
>> tag 66
>> Set clauses:
>> local-preference 200
>> origin igp
>> ip next-hop 192.0.2.1
>>
>> Static route:
>>
>> ip route vrf VPN_Internet 2.2.2.2 255.255.255.255 Null0 tag 66
>>
>> This works well the route 2.2.2.2 is advertised to the other routers
>> also the local-preference is applied. Butt the problem is that the ip
>> next-hop value is ignored. The next-hop address used is the ip
>> address of the advertising router.
>
> Aha, so you want to use this in an MPLS-*VPN* environment (three
> letters
> do make a difference :)
>
> You are right, the PE device in an MPLS-VPN will always do
> next-hop-self
> when it redistributes the routes into the vpnv4 mesh, this is how
> RFC2547bis works. You need to set the next-hop inbound or outbound at
> your vpnv4 neighbor. Please be aware that a MPLS-VPN PE looks into the
> global routing table to resolve the next-hop, so your route to
> 192.0.2.1
> needs to be in the global table..
>
> oli
More information about the cisco-nsp
mailing list