single AS versus multi AS MPLS VPN network

Hi,

I plan to have the following MPLS set up in our lab, it is a basic
interAS MPLS VPN set up, ip routing and MPLS seems to be working from PE to
PE but when I ping from CE to CE or ping vrf from PE to PE, it fails,
mpls debug messages enabled in all routers for traceroute vrf (PE to PE)
indicate that packet flow is ok to the remote PE & the final label that
the packet gets transferred to is the one that is associated with the
remote VPN route but ip lookup to reach the destination PE-CE network
seems to fail, the P router in the remote end indicates that host is
unreachable since TTL exceeded.

                   AS 100
AS200

CE1 --------- PE1
------------P-------------------------ASBR1--------------------ASBR2----------------------P-----------------PE2------------CE2
        fe                  fe            atm OC3
POS                       atm oc12             fe                   fe

      RIP      |                 OSPF                   |
eBGP                 |         OSPF                         |
RIP

I am using next-hop-self in the ASBRs and redistributing RIP routes
into mBGP in the PEs and redistributing BGP routes.
into RIP on both PEs (PE-CE).

Any help is appreciated,

The MPLS-OPS Archive

single AS versus multi AS MPLS VPN network