Good part about the topology is you can simulate whole topology using the logical routers and play with the various features or protocols. I am using my lab M7i router to simulate logical topology.

JUNIPER Certification book free PDF click here.

Topic you can cover with given topology:
1.   IGP
2.   MPLS
3.   BGP with RR
4.   MPLS VPN
5.   PE-CE: Static/OSPF/BGP
6.   Policy for SOO/VRF import export.
7.   Plain Multicast
8.   Internet connectivity with VRF & non-VRF interface
9.   MVPN with Draft Rosen.
10. IPv6

Routing instance used in topology: vpn-a

All routers loopback addresses start with 192.168.10X.X/32
X = router number

Primary and Backup LSPs:
Secondary Path will be established up front or not is depend on configuration. If secondary LSP is configured with the standby key word then it will be established before the primary LSP breaks. After primary path is recovered and primary LSP is established again, traffic will be automatically revert back to primary LSP and backup LSP will remain up for 60 seconds if it is not configured with the STANDBY keyword. To prevent traffic revert back, you can configure all LSP as a backup LSP as there is no mechanism of traffic revert back in case of the backup LSPs.

Adaptive keyword on secondary LSP trigger the sheared explicit behavior where primary LSP and backup LSP will share the same resources to establish the LSP.

Martini L2VPN Notes:
Do not for get to enable the LDP on interface. Draft Martini is based on LDP so when you use RSVP-TE then you will have to enable the ldp-tunneling, you will need RSVP-TE in both direction (Two tunnels) for bidirectional LDP communication and both tunnels will need ldp-tunneling. ldp-tunneling is targeted LDP sessions.

BGP between PEs:

Under BGP you will need inet unicast and inet-vpn address family.

lab@DEVANG-M7i# show protocols bgp
group ibgp {
    type internal;
    local-address 192.168.102.2;
    neighbor 192.168.104.4 {
        family inet {
            unicast;
        }
        family inet-vpn {
            unicast;
        }
    }
}

[edit logical-systems r2]


lab@LAB-M7i# show routing-instances vpn-a
instance-type vrf;
interface ge-0/3/1.12;
interface lo0.12;
route-distinguisher 1:1;
inactive: vrf-import vpna-import;
inactive: vrf-export ospf-to-bgp;
vrf-target target:1:1;
protocols {
    inactive: bgp {
        group ebgp {
            type external;
            import origin-import;
            export origin-export;
            as-override;
            neighbor 1.1.12.1 {
                peer-as 65001;
            }
        }
    }
    ospf {
        domain-id 192.168.102.2;
        export bgp-to-ospf;
        area 0.0.0.0 {
            interface ge-0/3/1.12;
            interface lo0.12;
        }
    }
    pim {
        vpn-group-address 239.3.3.3;
        rp {
            local {
                address 12.12.12.12;
            }
        }
        interface all {
            mode sparse;
        }
    }
}

[edit logical-systems r2]

[edit logical-systems r2]
lab@LAB-M7i# run show bgp summary
Groups: 1 Peers: 1 Down peers: 0
Table          Tot Paths  Act Paths Suppressed    History Damp State    Pending
inet.0                 1          1          0          0          0          0
inet6.0                0          0          0          0          0          0
Peer                     AS      InPkt     OutPkt    OutQ   Flaps Last Up/Dwn State|#Active/Received/Accepted/Damped...
192.16.100.2            100       5411       5402       0       0 1d 16:32:34 Establ
  inet.0: 1/1/1/0
  inet6.0: 0/0/0/0

Draft Rosen verification:

Default MDT: All PE has group of multicast address configured under the VRF, each VRF has unique MDT address that can be used to join the group at P-RP. In case of default MDT, all the PE participating in any particular MVPN will receive the multicast data whether they have receiver or not. If you have high bit stream multicast data and wide spread receiver and among those receiver only few receivers are interested in receiving the multicast data then use the Data MDT.

Data MDT: You can configure the group of mutlicast addresses that can be used to initiate the Data MDT. Data MDT can be triggered with configured threshold and threshold can be zero so as soon as the multicast source start sending the data, PE can initiate the Data MDT tunnel. To let the receiver end aware of Data MDT, source PE send out Data MDT join to standard all PIM router address 224.0.0.13. Once all receiver end PE receive the Data MDT join messages, interested receiver end PEs will join the Data MDT group.

In both the case Default or Data MDT, it will use the GRE encapsulations. Source IP will be the PE's loopback IP and destination IP will be the Default or Data MDT group address.

You should see the both end PE routers as a neighbor via "mt" interface. You can have default/data mdt for MVPN. You will need one Router in provider core acting as a P-RP (in our case R3) and one router in customer network or PE can act as a C-RP (in my lab I used PE2 router as a C-RP by configuring static RP under routing-instance vpn-a).

[edit logical-systems r2]
lab@LAB-M7i# run show pim neighbors logical-system r2 instance vpn-a
Instance: PIM.vpn-a
B = Bidirectional Capable, G = Generation Identifier,
H = Hello Option Holdtime, L = Hello Option LAN Prune Delay,
P = Hello Option DR Priority

Interface           IP V Mode        Option      Uptime Neighbor addr
ge-0/3/1.12          4 2             HPLG      00:31:26 1.1.12.1      
mt-1/2/0.37888       4 2             HPLG      00:26:37 45.45.45.45   

[edit logical-systems r2]
lab@LAB-M7i#

P-RP: