what do mean by "all available path"?
Do you mean available path to reach remote PE connected to remote CE. In that case your IGP will help you to find out the number of path.
The remore PE loopback can be reachable via multiple IGP path actuall protocol next-hop for remote CE routes.
Basically paths between PE-PE and CE-CE should be same untill CEs are multi-homed.
Hope these answers your query.
Thanks for your response.
I want to find all the path (primary and secondary) available in between CE in the SP. My requirement to show the customer the various path to their VPN services.
As SP you can easily find out number of paths available between two PEs. (Connected to two different CEs).
PE routing table will always give you the single best route (currently active) to reach remote PE loopback.
You can find out more in ospf/isis databases.
I am considering CE single homed and redundant paths available in SP backbone between PEs.
Definitely there will be single BGP session between PEs loopbacks but the reachability can be served by multiple IGP paths if available.
So basically your IGP database can help you.
Also you can check your TED if running RSVP-TE.
Hope these answeres your query.
I didnt got your pupose behind explaining these to your customer actually?
The purpose is to show the customer a graphical representation of the path in MPLS cloud for their VPN. Starting from the one site CPE to the other site CPE.
I can find the path by tracing the loopback IP of the PE from one end to other, but I want to know the other paths also in the cloud.
There is a tool in Cisco called Active Network Abastrct which will give the GUI of a VPN hop-by-hop. I want hop-by-hop to find manually.
Hope you are clear about my requirement.
If we need to find all the available paths for a remote CE from a local PE it will depend upon whether its a RR or non-RR design. If the MP-iBGP deisgn is non-RR the below vrf specific command
sh ip bgp vpnv4 vrf "vrf_name" will show us the MP-iBGP RT for that particular VPN. It will show us the next hop. Checking the route for same in the Global RT will show us the path(s) available for same ( load-balancing considered) .Then we can do a trace using the Local PE MP-iBGP loopback as source to remote PE's MP-iBGP loopback to get the physical Hops involved.
However if the design is RR-based there might be complications involved when the RR is in the forwarding path ie we have NHS being set to RR-MP-iBGP loopback and the trace using the Local PE MP-iBGP loopback as source to remote PE's MP-iBGP loopback will get us the physical Hops involved.
If we have redundant RRs being used with NHS being set then the output of sh ip bgp vpnv4 vrf "vrf_name" will show us two different available paths for the remote CE destination but just one being used.
RR-based design with no NHS being used will always to cater to single path for the remote CE detsination.
So in any case the actual path used for the remote CE connectivity would be a single unless we are using load-balancing.
Hope this helps you a bit on your requirement
Thanks & Regards