01-23-2004 01:37 PM - edited 03-02-2019 01:07 PM
In the lab, if my PE router doesn't support mpls ip command because it's a lower end router. can I still test vpn/vrf and bgp? In another words, can i update vrf routing table from router A to router B through BGP and both ends can talk each other without lable? I think it should be OK. Right?
01-23-2004 08:09 PM
You should indeed be able to get MP-BGP to exchange routes and VPN labels between the two PE routers. Unfortunately, that is as far as you will get since you can't setup an LSP between the two PEs.
Hope this helps,
01-25-2004 08:26 AM
Thanks.
By achieving that, I should be able to ping each other between these two vpn sites, right?
01-25-2004 04:16 PM
To do so, all of the provider-core routers will need to run BGP. This is a major advantage to MPLS - routers in the core do not need to be BGP aware.
Short answer, if all routers are running BGP, you will be able to ping between VPN sites.
HTH
Mark
01-25-2004 08:31 PM
You won't be able to get VPN connectivity if you don't have MPLS enabled in the core and that even if you have BGP running on all routers.
01-26-2004 12:02 PM
But, isn't mpls just a labeling mechnism? My understanding is same as Mark, I should have remote site's routing table in each site by running VPN/BGP even without MPLS.
01-26-2004 12:50 PM
You can't have MPLS/VPN without MPLS. What happens is your packet will have a VPN label assigned to it but as you try to forward it without an IGP label how will the network forward the packet from the source to the destination PE. Remember that this is not an IP packet anymore since you have a VPN label inserted.
01-26-2004 01:03 PM
Make sense. Now I see the point. How did VPN label get distributed or assigned? Is that from destination when update the routing table through BGP? Thanks
01-26-2004 01:55 PM
A VPN label is assigned to each prefix on the destination PE and then distributed via MP-BGP (AF vpnv4). This is done independantly from the LSP setup between the ingress and egress PEs.
Hope this helps,
03-17-2004 07:24 AM
Hi
Yeah, you are right
The idea is in the data (forwarding ) plane
The control plane ( MP-BGP VPN routes are propagated successfully but they are never utilized )
I tried this solution on my RPMs ( i have an MPLS cloud over ATM )
I tried also to configure GRE multipoint l3vpn OR L2TPv3 to provide a workaround but it seems that my RPMs don't support these configuartions
Thanks and best regards
Moustafa
01-25-2004 06:01 PM
Question on MPLS:
First question:global mode: mpls lable protocol ldp/tdp this statement is not supported on c2600-js-mz.122-6.bin version. however under interface : tag-switching ip is supported. Does that mean this version doesn't supprt ldp?
Second: how can I see the vpn lable, as I know there are two layer lables for each packet in VPN/BGP MPLS case.How is VPN label assigned to each packet? Thanks
01-25-2004 08:27 PM
1- LDP is not supported in 12.2. So what you get when enabling "tag-switching ip".
2- If you want to see what label is assigned to a remote VPN prefix, you can use the following command:
sh ip bgp vpnv4 all tags
You can also do a "sh ip cef vrf xxx
Hope this helps,
01-26-2004 11:58 AM
I thought show ip bgp vpnv4 all tags is showing the labels for mpls. Then, what command display labels for mpls, the first layer lable?
I used show mpls forwarding, but it shows the same information as show ip bgp vpnv4 all tags .
01-26-2004 12:39 PM
the "sh ip bgp vpn all tag" will display the VPN label (label learnt via vpnv4 address family).
If you want to see the IGP label (label used to get from one PE to another) then you need to do a
sh ip cef vrf
which should show you both the VPN and IGP labels or
sh ip cef
Which will show the IGP label.
Hope this helps,
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: