RSVP Signaled Point-to-Multipoint LSPs

In This Chapter

This section provides information about RSVP signaled point-to-multipoint LSPs.

Topics in this section include:

•	Applicability

•

Overview

•	Configuration

•	Conclusion

Applicability

This feature is applicable to all of the 7750 and 7710 SR series (except SR-1). Tested on release 7.0R5. On all nodes involved with the LSP, at least chassis mode C is required. This means that modular systems should be equipped with IOM-2 linecards or higher. This is supported on 7450 ESS-7 or ESS-12 in mixed-mode since 8.0R1. The 7750 SR-c4 is supported from 8.0R4 and higher.

Overview

Point-to-MultiPoint (P2MP) MPLS label switched path (LSP) allows the source of multicast traffic to forward packets to one or many multicast receivers over a network without requiring a multicast protocol, such as PIM, to be configured in the network. A P2MP LSP tree is established in the control plane which path consists of a head-end node, one or many branch and BUD nodes, and the leaf nodes. Packets injected by the head-end node are replicated in the data plane at the branching nodes before they are delivered to the leaf nodes.

Similar to point-to-point (P2P) LSPs, also P2MP LSPs are unidirectional, originating on a head-end node (the ingress LER) and terminating on one or more leaf node(s) (the egress LER(s)). Initially, RSVP is used as signaling protocol. A P2MP LSP is modeled as a set of root-to-leaf sub LSPs (S2L). Each S2L is modeled as a point-to-point LSP in the control plane. This means that each S2L has it’s own PATH/RESV messages. This is called the de-aggregated method.

The forwarding of multicast packets to the LSP tree is based on static multicast routes initially but will evolve to BGP based VPN routes in the future. Forwarding multicast packets is initially done over P2MP RSVP LSPs in the base router instance but will evolve to VPRNs.

RSVP signalled P2MP LSPs can have fast reroute (FRR) enabled, the facility method (one-to-many) with link protection is supported.

 

Figure 169: P2MP Network Topology

Configuration

The following sections describe the tasks you must perform to configure RSVP signaled point-to-multipoint LSPs.

 

Configuring the IP/MPLS Network

After configuring the cards and MDAs, the correct chassis-mode must be set (C or higher) on all MPLS nodes along the P2MP LSP.

A:PE-[1..7]# configure system chassis-mode c

A:PE-[1..7]# show chassis | match chassis mode

    Admin chassis mode            : c

    Oper chassis mode             : c

The system addresses and L3 interface addresses are configured according to Figure 169. An interior gateway protocol (IGP) is needed to distribute routing information on all PEs. In our case, the IGP is OSPF using the backbone area (area 0.0.0.0). A configuration example is shown for PE-1. A similar configuration is needed on all 7 PEs

A:PE-1# configure router 

        interface "int-PE-1-PE-2"

            address 192.168.12.1/30

            port 1/1/3

        exit

        interface "int-PE-1-PE-3"

            address 192.168.13.1/30

            port 1/1/2

        exit

        interface "system"

            address 192.0.2.1/32

        exit

A:PE-1# configure router 

        ospf

            traffic-engineering

            area 0.0.0.0

                interface "system"

                exit                  

                interface "int-PE-1-PE-3"

                    interface-type point-to-point

                exit                  

                interface "int-PE-1-PE-2"

                    interface-type point-to-point

                exit                  

            exit                      

        exit

Since fast reroute (FRR) will be enabled on the P2MP LSP, traffic engineering (TE) is needed on the IGP. By doing this, OSPF will generate opaque LSAs which are collected in a traffic engineering database (TED), separate from the traditional OSPF topology database. OSPF interfaces are setup as type ‘point-to-point’ to improve convergence, no DR/BDR election process is done.¹

In our example on PE-1, the interface towards the multicast source is configured as an IES service. This could have been on a router interface instead.

Similar IES services are configured on PE-7 and PE-6 for multicast client 1 and multicast client 2.

A:PE-1 # configure service 

         …

         ies 1 customer 1 create

            interface "to-MC-source" create

                address 192.168.9.1/30

                sap 1/1/1 create

                exit

            exit

            no shutdown

         exit

A:PE-7# configure service 

        …

        ies 1 customer 1 create

            interface "to-MC-client1" create

                address 192.168.10.1/30

                sap 1/1/6 create

                exit

            exit

            no shutdown

        exit

A:PE-6# configure service 

        …

        ies 1 customer 1 create

            interface "to-MC-client2" create

                address 192.168.11.1/30

                sap 1/1/6 create

                exit

            exit

            no shutdown

        exit

 

 

 

 

To verify that OSPF neighbors are up (state Full), show router ospf neighbor is performed. To check if L3 interface addresses/subnets are known on all PEs, show router route-table or show router fib iom-card-slot will display the content of the forwarding information base (FIB).

A:PE-1# show router ospf neighbor 

===============================================================================

OSPF Neighbors

===============================================================================

Interface-Name                   Rtr Id          State      Pri  RetxQ   TTL   

-------------------------------------------------------------------------------

int-PE-1-PE-3                    192.0.2.3       Full       1    0       36    

int-PE-1-PE-2                    192.0.2.2       Full       1    0       32    

-------------------------------------------------------------------------------

No. of Neighbors: 2

===============================================================================

A:PE-1# show router route-table 

===============================================================================

Route Table (Router: Base)

===============================================================================

Dest Prefix                                   Type    Proto    Age         Pref

       Next Hop[Interface Name]                                     Metric     

-------------------------------------------------------------------------------

192.0.2.1/32                                  Local   Local    15d04h27m   0   

       system                                                       0

192.0.2.2/32                                  Remote  OSPF     01d22h28m   10  

       192.168.12.2                                                 10000

192.0.2.3/32                                  Remote  OSPF     01d22h27m   10  

       192.168.13.2                                                 1000

192.0.2.4/32                                  Remote  OSPF     01d22h24m   10  

       192.168.13.2                                                 3000

192.0.2.5/32                                  Remote  OSPF     01d22h24m   10  

       192.168.13.2                                                 2000

192.0.2.6/32                                  Remote  OSPF     01d22h23m   10  

       192.168.13.2                                                 4000

192.0.2.7/32                                  Remote  OSPF     01d22h22m   10  

       192.168.13.2                                                 5000

192.168.12.0/30                               Local   Local    01d22h31m   0   

       int-PE-1-PE-2                                                0

192.168.13.0/30                               Local   Local    01d22h31m   0   

       int-PE-1-PE-3                                                0

192.168.23.0/30                               Remote  OSPF     01d22h27m   10  

       192.168.13.2                                                 11000

192.168.24.0/30                               Remote  OSPF     01d22h24m   10  

       192.168.13.2                                                 13000

192.168.35.0/30                               Remote  OSPF     01d22h26m   10  

       192.168.13.2                                                 2000

192.168.45.0/30                               Remote  OSPF     01d22h24m   10  

       192.168.13.2                                                 3000

192.168.46.0/30                               Remote  OSPF     01d22h24m   10  

       192.168.13.2                                                 4000

192.168.57.0/30                               Remote  OSPF     01d22h24m   10  

       192.168.13.2                                                 12000

192.168.67.0/30                               Remote  OSPF     01d22h23m   10  

       192.168.13.2                                                 5000

-------------------------------------------------------------------------------

No. of Routes: 16

===============================================================================

A:PE-1# show router fib 1 

===============================================================================

FIB Display

===============================================================================

Prefix                                                      Protocol           

    NextHop                                                                    

-------------------------------------------------------------------------------

192.0.2.1/32                                                LOCAL              

    192.0.2.1 (system)

192.0.2.2/32                                                OSPF               

    192.168.12.2 (int-PE-1-PE-2)

192.0.2.3/32                                                OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.0.2.4/32                                                OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.0.2.5/32                                                OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.0.2.6/32                                                OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.0.2.7/32                                                OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.168.12.0/30                                             LOCAL              

    192.168.12.0 (int-PE-1-PE-2)

192.168.13.0/30                                             LOCAL              

    192.168.13.0 (int-PE-1-PE-3)

192.168.23.0/30                                             OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.168.24.0/30                                             OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.168.35.0/30                                             OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.168.45.0/30                                             OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.168.46.0/30                                             OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.168.57.0/30                                             OSPF               

    192.168.13.2 (int-PE-1-PE-3)

192.168.67.0/30                                             OSPF               

    192.168.13.2 (int-PE-1-PE-3)

-------------------------------------------------------------------------------

Total Entries : 16

-------------------------------------------------------------------------------

===============================================================================

 

The next step in the process of setting up a P2MP LSP, is enabling our L3 interfaces in the MPLS and RSVP context on all involved PE nodes (PE-1 <=> PE-7). By default, the system interface is put automatically within the MPLS/RSVP context. When an interface is put in the MPLS context, 7x50 copies it also in the RSVP context. Explicit enabling of MPLS and RSVP context is done by the no shutdown command. Below you can find the MPLS/RSVP configuration for PE-1.

A:PE-1# configure router mpls no shutdown

A:PE-1# configure router rsvp no shutdown

A:PE-1# configure router mpls

            interface "system"

            exit

            interface "int-PE-1-PE-3"

            exit

            interface "int-PE-1-PE-2"

            exit

            no shutdown

A:PE-1# configure router rsvp

            interface "system"

            exit

            interface "int-PE-1-PE-3"

            exit

            interface "int-PE-1-PE-2"

            exit

            no shutdown

Figure 170: P2MP LSP LSP-p2mp-1

 

Configuring P2MP RSVP LSP

See Figure 170 for a description of configuring PSMP LSPs.

A P2MP LSP (LSP-p2mp-1) will be setup from PE-1 acting as head-end node and PE-6 and PE-7 acting as leaf nodes. Because FRR is enabled, Constrained Shortest Path First (CSPF) is enabled to do route calculations on the traffic engineering database (TED). FRR method facility is used without node protection, facility stands for one-to-many meaning that one bypass tunnel can protect a set of primary LSPs with similar backup constraints. When a link failure occurs on one of the active S2L paths, the Point of Local Repair (PLR) node will push an additional MPLS label on the incoming MPLS packet before sending it into the bypass tunnel downstream towards the merge point (MP) node.

In the first example, OSPF (our IGP) will do the path calculation to the two destinations (PE-6 and PE-7). The intermediate hops of the LSP are dynamically assigned by OSPF best route selection, thus S2L paths follows the IGP least cost path. For this, an MPLS path loose is configured without specifying any strict/loose hops.

A:PE-1# configure router mpls 

            …

            path "loose"

                no shutdown

            exit

Creation of the P2MP LSP itself is done on the ingress LER or head-end node (PE-1 in our example) and can be seen in following CLI output. P2MP name is LSP-p2mp-1. A create time keyword p2mp-lsp is added in addition to the P2MP name to make a distinction in configuration between normal point-to-point LSPs and point-to-multipoint LSPs. A primary P2MP instance is initiated using the primary-p2mp-instance keyword accompanied with the P2MP instance name p-LSP-p2mp-1. Within this primary P2MP instance, the different S2Ls are defined using the s2l-path keyword. Be aware that the same MPLS path name can be used for different S2Ls as long as the destination is different (to command).

A:PE-1# configure router mpls 

            …

            lsp "LSP-p2mp-1" p2mp-lsp

                cspf

                fast-reroute facility

                    no node-protect

                exit

                primary-p2mp-instance "p-LSP-p2mp-1"

                    s2l-path "loose" to 192.0.2.6

                    exit

                    s2l-path "loose" to 192.0.2.7

                    exit

                exit

                no shutdown

            exit

            no shutdown

On the head-end LER node of the P2MP LSP, several show commands can be used. First set of show commands are used to verify the administrative and operational state of the P2MP LSP and his different S2L paths (including FRR bypass information). In our example, ‘LSP-p2mp-1’ P2MP LSP has two active S2L paths towards leaf node PE-6 and leaf node PE-7

A:PE-1# show router mpls p2mp-lsp 

===============================================================================

MPLS P2MP LSPs

===============================================================================

LSP Name                           P2mp-ID             Fastfail     Adm   Opr  

                                                       Config                  

-------------------------------------------------------------------------------

LSP-p2mp-1                          0                   Yes          Up    Up   

-------------------------------------------------------------------------------

LSPs : 1

===============================================================================

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-1 detail 

===============================================================================

MPLS P2MP LSPs (Originating) (Detail)

===============================================================================

-------------------------------------------------------------------------------

Type : Originating 

-------------------------------------------------------------------------------

LSP Name    : LSP-p2mp-1                       LSP Tunnel ID  : 11

From        : 192.0.2.1                        To             : 0.0.0.0

Adm State   : Up                               Oper State     : Up

LSP Up Time : 0d 00:08:20                      LSP Down Time  : 0d 00:00:00

Transitions : 1                                Path Changes   : 1

Retry Limit : 0                                Retry Timer    : 30 sec

Signaling   : RSVP                             Resv. Style    : SE

Hop Limit   : 255                              Negotiated MTU : n/a

Adaptive    : Enabled                          ClassType      : 0

FastReroute : Enabled                          Oper FR        : Enabled

FR Method   : Facility                         FR Hop Limit   : 16

FR Bandwidth: 0 Mbps                           FR Node Protect: Disabled

FR Object   : Enabled                          

CSPF        : Enabled                          ADSPEC         : Disabled

Metric      : Disabled                         Use TE metric  : Disabled

Include Grps:                                  Exclude Grps   :  

None                                           None

Type        : P2mpLsp                          Least Fill     : Disabled

LdpOverRsvp : Disabled                         VprnAutoBind   : Disabled

Oper Metric : Disabled                         

P2MPInstance: p-LSP-p2mp-1                     P2MP-Inst-type : Primary

S2l-Name    : loose                            To             : 192.0.2.6

S2l-Name    : loose                            To             : 192.0.2.7

===============================================================================

A:PE-1# show router mpls p2mp-info 

===============================================================================

MPLS P2MP Cross Connect Information

===============================================================================

-------------------------------------------------------------------------------

Source IP Address    : 192.0.2.1             Tunnel ID     : 11

P2MP ID              : 0                     Lsp ID        : 27648

S2L Name             : LSP-p2mp-1::loose     To            : 192.0.2.6

Out Interface        : 1/1/2                 Out Label     : 131068

Num. of S2ls         : 2                     

-------------------------------------------------------------------------------

S2L LSP-p2mp-1::loose

-------------------------------------------------------------------------------

Source IP Address    : 192.0.2.1             Tunnel ID     : 11

P2MP ID              : 0                     Lsp ID        : 27648

S2L Name             : LSP-p2mp-1::loose     To            : 192.0.2.7

Out Interface        : 1/1/2                 Out Label     : 131068

Num. of S2ls         : 2                     

-------------------------------------------------------------------------------

P2MP Cross-connect instances : 2

===============================================================================

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-1 p2mp-instance p-LSP-p2mp-1 

===============================================================================

MPLS P2MP Instance (Originating)

===============================================================================

-------------------------------------------------------------------------------

Type : Originating 

-------------------------------------------------------------------------------

LSP Name    : LSP-p2mp-1                       LSP Tunnel ID  : 11

P2MP ID     : 0                                

Adm State   : Up                               Oper State     : Up

P2MPInstance: p-LSP-p2mp-1                     P2MP-Inst-type : Primary

P2MP Inst Id: 3                                P2MP Lsp Id    : 27648

Inst Admin  : Up                               Inst Oper      : Up

Inst Up Time: 0d 00:12:14                      Inst Dn Time   : 0d 00:00:00

Hop Limit   : 255                              Adaptive       : Enabled

Record Route: Record                           Record Label   : Record

Include Grps:                                  Exclude Grps   :  

None                                           None

Bandwidth   : No Reservation                   Oper Bw        : 0 Mbps

S2l-Name    : loose                            To             : 192.0.2.6

S2l Admin   : Up                               S2l Oper       : Up

S2l-Name    : loose                            To             : 192.0.2.7

S2l Admin   : Up                               S2l Oper       : Up

-------------------------------------------------------------------------------

P2MP instances : 1

===============================================================================

 

 

 

 

FRR information can be displayed in detail for each S2L path. From this moment onward, we will only focus on the S2L path towards PE-7. As you can see in the show command, link protection is present for links PE-1 <=> PE-3, PE-3 <=> PE-5 and PE-5 <=> PE-7 (‘@’-reference inside show command).

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-1 p2mp-instance p-LSP-p2mp-1 s2l loose to 192.0.2.7 detail 

===============================================================================

MPLS LSP LSP-p2mp-1 S2L loose (Detail)

===============================================================================

Legend : 

    @ - Detour Available                          # - Detour In Use

    b - Bandwidth Protected                       n - Node Protected

    s - Soft Preemption                   

===============================================================================

-------------------------------------------------------------------------------

LSP LSP-p2mp-1 S2L loose

-------------------------------------------------------------------------------

LSP Name    : LSP-p2mp-1                         S2l LSP ID  : 27652

P2MP ID     : 0                                  S2l Grp Id  : 2

Adm State   : Up                                 Oper State  : Up

S2l State:  : Active                                         :  

S2L Name    : loose                              To          : 192.0.2.7

S2l Admin   : Up                                 S2l Oper    : Up

OutInterface: 1/1/2                              Out Label   : 131071

S2L Up Time : 0d 00:26:40                        S2L Dn Time : 0d 00:00:00

RetryAttempt: 0                                  NextRetryIn : 0 sec

S2L Trans   : 3                                  CSPF Queries: 3

Failure Code: noError                            Failure Node: n/a

ExplicitHops:                                    

    No Hops Specified

Actual Hops :                                    

    192.168.13.1(192.0.2.1) @                    Record Label    : N/A

 -> 192.168.13.2(192.0.2.3) @                    Record Label    : 131071

 -> 192.168.35.2(192.0.2.5) @                    Record Label    : 131071

 -> 192.168.57.2(192.0.2.7)                      Record Label    : 131068

ComputedHops:                                    

    192.168.13.1    -> 192.168.13.2    -> 192.168.35.2    -> 192.168.57.2   

LastResignal: n/a                                

===============================================================================

More in detail, show router mpls bypass-tunnel can be used. Actual Hops gives you the explicit hops of the bypass tunnel used to avoid link PE-1 <=> PE-3. On PE-1 node the MPLS path PE-1 <=> PE-2 <=> PE-3 is followed (see also Figure 170).

A:PE-1# show router mpls bypass-tunnel protected-lsp p2mp detail 

===============================================================================

MPLS Bypass Tunnels (Detail)

===============================================================================

-------------------------------------------------------------------------------

bypass-link192.168.13.2

-------------------------------------------------------------------------------

To              : 192.168.23.2       State               : Up

Out I/F         : 1/1/3              Out Label           : 131064

Up Time         : 1d 23:25:14        Active Time         : n/a

Reserved BW     : 0 Kbps             Protected LSP Count : 3

Type            : P2mp               

SetupPriority   : 7                  Hold Priority       : 0

Class Type      : 0                  

Actual Hops     :                    

    192.168.12.1    -> 192.168.12.2    -> 192.168.23.2   

Protected LSPs -

LSP Name        : LSP-p2mp-1::loose

From            : 192.0.2.1          To                  : 192.0.2.7

Avoid Node/Hop  : 192.168.13.2       Downstream Label    : 131071

Bandwidth       : 0 Kbps

On PE-3 node the MPLS path PE-3 <=> PE-2 <=> PE-4 <=> PE-5 is followed (see Figure 170) to avoid link PE-3 <=> PE-5.

A:PE-3# show router mpls bypass-tunnel protected-lsp p2mp detail 

===============================================================================

MPLS Bypass Tunnels (Detail)

===============================================================================

-------------------------------------------------------------------------------

bypass-link192.168.35.2

-------------------------------------------------------------------------------

To              : 192.168.45.2       State               : Up

Out I/F         : 1/1/4              Out Label           : 131071

Up Time         : 1d 23:28:31        Active Time         : n/a

Reserved BW     : 0 Kbps             Protected LSP Count : 3

Type            : P2mp               

SetupPriority   : 7                  Hold Priority       : 0

Class Type      : 0                  

Actual Hops     :                    

    192.168.23.2    -> 192.168.23.1    -> 192.168.24.2    -> 192.168.45.2   

Protected LSPs -

LSP Name        : LSP-p2mp-1::loose

From            : 192.0.2.1          To                  : 192.0.2.7

Avoid Node/Hop  : 192.168.35.2       Downstream Label    : 131071

Bandwidth       : 0 Kbps             

===============================================================================

A similar output can be seen on PE-5 node also. The MPLS path PE-5 <=> PE-4 <=> PE-6 <=> PE-7 is followed (see also Figure 170).

On the transit LSRs and egress LER/leaf node (see also Figure 170), the show router mpls p2mp-info command can be used. Attached is the show command on PE-3 node included for S2L path to 192.0.2.7. Similar outputs are possible for PE-5 and PE-7 node.

A:PE-3# show router mpls p2mp-info type transit 

===============================================================================

MPLS P2MP LSPs (Transit)

===============================================================================

-------------------------------------------------------------------------------

S2L LSP-p2mp-1::loose

-------------------------------------------------------------------------------

Source IP Address    : 192.0.2.1             Tunnel ID     : 11

P2MP ID              : 0                     Lsp ID        : 27652

S2L Name             : LSP-p2mp-1::loose     To            : 192.0.2.7

Out Interface        : 1/1/1                 Out Label     : 131071

Num. of S2ls         : 1                     

-------------------------------------------------------------------------------

P2MP Cross-connect instances : 1

===============================================================================

 

Mapping Multicast Traffic

To map multicast traffic into the LSP tree from the head-end node until leaf node, PIM and IGMP configurations are needed on the head-end node (PE-1) and leaf nodes (PE-6 and PE-7) of the P2MP RSVP LSP. The intermediate nodes (transit LSR, branch LSR and BUD LSR) do not need any explicit configuration for that.

 

Head-end Node (Ingress LER)

PIM must be enabled on the interface towards the MC source and PIM must be enabled on the tunnel interface. A tunnel interface should be seen as an internal representation of a specific P2MP LSP. Creation is done within the PIM context using the tunnel-interface rsvp-p2mp command followed by the P2MP LSP name. Translated into configuration commands, this becomes :

A:PE-1# configure router pim 

            …

            interface "to-MC-source"

            exit

            tunnel-interface rsvp-p2mp LSP-p2mp-1 

            …

In the data path, when a multicast packet is received on an interface a successful Reverse Path Forwarding (RPF) check must be done for the source address otherwise the packet will be dropped.

Besides enabling PIM on the tunnel interface also IGMP is enabled to do a static <S,G> or <*,G> join of a multicast group address (227.1.1.1 in our example) to the tunnel interface/P2MP LSP. Be aware that there is always a one-to-one mapping between <S,G> or <*,G> and a tunnel interface/P2MP LSP. In our example a < S,G > will be configured. A <*,G> join scenario is included in Additional Topics .

A:PE-1# configure router igmp 

        …

        tunnel-interface rsvp-p2mp "LSP-p2mp-1" 

            static

                group 227.1.1.1

                    source 192.168.9.2

                exit

            exit

        exit

 

 

 

The show router pim tunnel-interface command shows you the admin state of the tunnel interface and an association to an internal local ifindex (73730 in the example).

A:PE-1# show router pim tunnel-interface 

===============================================================================

PIM Tunnel-Interfaces

===============================================================================

LSP                                     Sender Addr         IfIndex  Oper     

-------------------------------------------------------------------------------

LSP-p2mp-1                              None                73730     Up       

-------------------------------------------------------------------------------

Interfaces : 2

===============================================================================

With show router igmp group you see the configured <S,G> entry and outgoing interface (= tunnel interface), represented by mpls-if-73730.

A:PE-1# show router igmp group 227.1.1.1 

===============================================================================

IGMP Groups

===============================================================================

(192.168.9.2,227.1.1.1)                Up Time : 0d 00:06:27

    Fwd List  : mpls-if-73730 

-------------------------------------------------------------------------------

(*,G)/(S,G) Entries : 1

===============================================================================

At this moment in time, users can verify if multicast traffic is using P2MP LSP at the head-end node using the show router pim group group-address detail command

A:PE-1# show router pim group 227.1.1.1 detail 

===============================================================================

PIM Source Group ipv4

===============================================================================

Group Address      : 227.1.1.1

Source Address     : 192.168.9.2

RP Address         : 192.0.2.1

Flags              : spt                Type               : (S,G)

MRIB Next Hop      : 192.168.9.2

MRIB Src Flags     : direct             Keepalive Timer Exp: 0d 00:03:29

Up Time            : 0d 00:08:59        Resolved By        : rtable-u

Up JP State        : Joined             Up JP Expiry       : 0d 00:00:00

Up JP Rpt          : Not Joined StarG   Up JP Rpt Override : 0d 00:00:00

Register State     : Join               Register Stop Exp  : 0d 00:00:00

Reg From Anycast RP: No                 

Rpf Neighbor       : 192.168.9.2

Incoming Intf      : to-MC-source

Outgoing Intf List : mpls-if-73730

Curr Fwding Rate   : 14.7 kbps          

Forwarded Packets  : 649                Discarded Packets  : 0

Forwarded Octets   : 29854              RPF Mismatches     : 0

Spt threshold      : 0 kbps             ECMP opt threshold : 7

Admin bandwidth    : 1 kbps             

-------------------------------------------------------------------------------

Groups : 1

===============================================================================

 

Leaf Node (Egress LER)

In the PIM context, the same tunnel interface must be created as the head-end node with in addition an explicit reference to the head-end system address, translated into the sender systemIP_head-end_node parameter.

A:PE-7# configure router pim 

            tunnel-interface rsvp-p2mp LSP-p2mp-1 sender 192.0.2.1

The show router pim tunnel-interface command shows you the admin state of the tunnel interface and an association to an internal local ifindex (73730 in our example, by coincidence the same ifindex as the one on the head-end node/PE-1).

A:PE-7# show router pim tunnel-interface 

===============================================================================

PIM Tunnel-Interfaces

===============================================================================

LSP                                     Sender Addr         If Index  Oper     

-------------------------------------------------------------------------------

LSP-p2mp-1                              192.0.2.1           73730     Up       

-------------------------------------------------------------------------------

Interfaces : 1

===============================================================================

The main goal on the leaf node(s) is to get traffic off the P2MP LSP/tunnel interface. This is done using a multicast information policy (multicast-info-policy). Inside this MC policy, a range of multicast group addresses must be defined under a bundle context (bundle1) in order to see traffic (channel) . Also inside the bundle context, the P2MP LSP is presented by the tunnel interface (primary-tunnel-interface). Translated into configuration commands, this becomes :

A:PE-7# configure mcast-management 

            multicast-info-policy "p2mp-pol" create

                bundle "bundle1" create

                    primary-tunnel-interface rsvp-p2mp LSP-p2mp-1 sender 192.0.2.1

                    channel "227.1.1.1" "227.1.1.1" create

                    exit

                exit

                bundle "default" create

                exit

            exit

 

Note: The channel command must be seen as a range command with a start-mc-group-address and an end-mc-group-address. In our example, only one MC group address, 227.1.1.1 is seen.

The configured multicast information policy must be applied to the base router instance.

A:PE-7# configure router multicast-info-policy p2mp-pol

On the leaf node (PE-7/PE-6), MC clients are connected. IGMP is enabled on those MC clients with a static <S,G> join to redirect MC traffic downstream to the MC client. Translated into configuration commands, this becomes :

A:PE-7 # configure router igmp 

        …

        interface "to-MC-client1"

            static

                group 227.1.1.1

                    source 192.168.9.2

                exit

            exit

        exit

 

With show router igmp group you see the configured <S,G> entry and outgoing interface (= to-MC-client1).

A:PE-7# show router igmp group 227.1.1.1 

===============================================================================

IGMP Groups

===============================================================================

(192.168.9.2,227.1.1.1)                Up Time : 0d 00:00:43

    Fwd List  : to-MC-client1 

-------------------------------------------------------------------------------

(*,G)/(S,G) Entries : 1

===============================================================================

Now, users can verify if multicast traffic is sent to MC client using the show router pim group group-address detail command

A:PE-7# show router pim group 227.1.1.1 detail 

===============================================================================

PIM Source Group ipv4

===============================================================================

Group Address      : 227.1.1.1

Source Address     : 192.168.9.2

RP Address         : 0

Flags              :                    Type               : (S,G)

MRIB Next Hop      : 

MRIB Src Flags     : remote             Keepalive Timer    : Not Running

Up Time            : 0d 00:27:43        Resolved By        : unresolved

Up JP State        : Joined             Up JP Expiry       : 0d 00:00:17

Up JP Rpt          : Not Joined StarG   Up JP Rpt Override : 0d 00:00:00

Register State     : No Info            

Reg From Anycast RP: No                 

Rpf Neighbor       : 

Incoming Intf      : mpls-if-73730

Outgoing Intf List : to-MC-client1

Curr Fwding Rate   : 14.7 kbps          

Forwarded Packets  : 66340              Discarded Packets  : 0

Forwarded Octets   : 3051640            RPF Mismatches     : 0

Spt threshold      : 0 kbps             ECMP opt threshold : 7

Admin bandwidth    : 1 kbps             

-------------------------------------------------------------------------------

Groups : 1

===============================================================================

 

OAM Tool

P2P LSP operation and maintenance (OAM) commands (oam lsp-ping and oam lsp-trace) are extended for P2MP LSP. The user can instruct the head-end node to generate an P2MP LSP ping or a P2MP LSP trace by entering the command oam p2mp-lsp-ping or oam p2mp-lsp-trace. The P2MP OAM extensions are defined in draft-ietf-mpls-p2mp-lsp-ping.

For P2MP LSP ping, the echo request is sent on the active P2MP instance and is replicated in the data path over all branches of the P2MP LSP instance. By default, all egress LER nodes which are leaves of the P2MP LSP instance will reply. Echo reply messages can be reduced by configuring the s2l-dest-address (a maximum of up to five egress nodes in a single run of the OAM command). Replies are sent by IP.

A:PE-1# oam p2mp-lsp-ping 

  - p2mp-lsp-ping <lsp-name> [p2mp-instance <instance-name> [s2l-dest-address

    <ip-address> [...(upto 5 max)]]] [fc <fc-name> [profile {in|out}]] [size

    <octets>] [ttl <label-ttl>] [timeout <timeout>] [detail]

 <lsp-name>           : [32 chars max]

 <instance-name>      : [32 chars max]

 <ip-address>         : ipv4 address    a.b.c.d - Up to 5 addresses permitted

 <profile>            : in|out - Default: out

 <fc-name>            : be|l2|af|l1|h2|ef|h1|nc - Default: be

 <octets>             : [92|97..9198]- Default: 92

 <label-ttl>          : [1..255] - Default: 255

 <timeout>            : [1..120] seconds - Default: 10

 <detail>             : keyword - displays detailed information

A:PE-1# oam p2mp-lsp-ping LSP-p2mp-1 detail 

P2MP LSP LSP-p2mp-1: 92 bytes MPLS payload

===============================================================================

S2L Information

===============================================================================

From             RTT                  Return Code

-------------------------------------------------------------------------------

192.0.2.6        =3.35ms              EgressRtr(3)

192.0.2.7        =6.14ms              EgressRtr(3)

===============================================================================

Total S2L configured/up/responded = 2/2/2,

          round-trip min/avg/max  = 3.35 / 4.74 / 6.14 ms

Responses based on return code:

        EgressRtr(3)=2

Note: Return codes are based on RFC4379, with value 3 => replying router is an egress for the FEC at stack-depth.

P2MP LSP trace allows the user to trace the path of a single S2L path of a P2MP LSP from head-end node to leaf node. By the use of the downstream mapping TLV, each node along the S2L path can fill in the appropriate flags : B or E flag. B-flag is set when responding node is a branch LSR, E-flag is set when responding node is an egress LER.

A:PE-1# oam p2mp-lsp-trace           

  - p2mp-lsp-trace <lsp-name> p2mp-instance <instance-name> s2l-dest-address

    <ip-address> [fc <fc-name> [profile {in|out}]] [size <octets>] [max-fail

    <no-response-count>] [probe-count <probes-per-hop>] [min-ttl

    <min-label-ttl>] [max-ttl <max-label-ttl>] [timeout <timeout>]  [interval

    <interval>] [detail]

 <lsp-name>           : [32 chars max]

 <instance-name>      : [32 chars max]

 <ip-address>         : ipv4 address    a.b.c.d

 <fc-name>            : be|l2|af|l1|h2|ef|h1|nc - Default: be

 <profile>            : in|out - Default: out

 <octets>             : [128..9198] - Default: 128

 <no-response-count>  : [1..10] - Default: 5

 <probes-per-hop>     : [1..10] - Default: 1

 <min-label-ttl>      : [1..255] - Default: 1

 <max-label-ttl>      : [1..255] - Default: 30

 <timeout>            : [1..60] seconds

 <detail>             : keyword - displays detailed information

A:PE-1# oam p2mp-lsp-trace LSP-p2mp-1 p2mp-instance p-LSP-p2mp-1 s2l-dest-address 192.0.2.7 detail 

P2MP LSP LSP-p2mp-1: 128 bytes MPLS payload

P2MP Instance p-LSP-p2mp-1, S2L Egress 192.0.2.7

  1  192.0.2.3  rtt=4.87 ms rc=8(DSRtrMatchLabel)

     DS 1: IfAddr 192.168.35.2 MRU=1500 label=131064 proto=4(RSVP-TE) B/E flags:0/0 

  2  192.0.2.5  rtt=8.02 ms rc=8(DSRtrMatchLabel)

     DS 1: IfAddr 192.168.57.2 MRU=1500 label=131066 proto=4(RSVP-TE) B/E flags:0/0 

  3  192.0.2.7  rtt=4.44 ms rc=3(EgressRtr)

Note: Return codes are based on RFC4379, with value ‘8’ => label switched at stack-depth. Meaning it is an transit LSR, doing MPLS label swapping. No B or E flag is set.

 

Additional Topics

<*,G> IGMP join iso <S,G> IGMP join

In the Head-end Node (Ingress LER) and Leaf Node (Egress LER) steps, a source specific IGMP join (<S,G> join) was used at the head-end node and leaf nodes. Another possibility is to use a source unknown or starg IGMP join (<*,G> join). When doing the latter, a rendezvous point (RP) must be defined in the PIM network. The RP allows multicast data flows between sources and receivers to meet at a predefined network location (in our example, loopback address of PE-1 node). It must be seen as an intermediate device to establish a multicast flow.

RP can be defined in a dynamic way (BSR protocol) or a static way. In our example we choose the static way meaning that on all involved PIM nodes, the RP address will be statically configured. The following output shows translated in configuration commands.

A:PE-1/PE-6/PE-7# configure router pim 

            …

            rp

                static

                    address 192.0.2.1

                        group-prefix 227.1.1.1/32

                    exit

                exit

                …

            exit

 

The group-prefix a mandatory keyword is specified and references a group address or group address range for which this rendez-vous point will be used.

A:PE-1/PE-6/PE-7# show router pim rp 

===============================================================================

PIM RP Set ipv4

===============================================================================

Group Address                                                Hold Expiry       

  RP Address                                   Type     Prio Time Time         

-------------------------------------------------------------------------------

227.1.1.1/32

  192.0.2.1                                    Static   1    N/A  N/A          

-------------------------------------------------------------------------------

Group Prefixes : 1

===============================================================================

As previously mentioned, the configuration of the <*,G> IGMP join is done on the head-end node (PE-1) and leaf nodes (PE-6 and PE-7)

A:PE-1# configure router igmp 

        tunnel-interface rsvp-p2mp "p-to-mp-1" 

            static

                group 227.1.1.1

                    starg

                exit

            exit

        exit

A:PE-6# configure router igmp 

        interface "to-MC-client2"

            static

                group 227.1.1.1

                    starg

                exit

            exit

        exit

A:PE-7# configure router igmp 

        interface "to-MC-client1"

            static

                group 227.1.1.1

                    starg

                exit

            exit

        exit

The same previous show command can be used to verify the multicast traffic on head-end node and leaf nodes, show router igmp group 227.1.1.1 and show router pim group 227.1.1.1 detail.

A:PE-7# show router pim group 227.1.1.1 detail 

===============================================================================

PIM Source Group ipv4

===============================================================================

Group Address      : 227.1.1.1

Source Address     : *

RP Address         : 192.0.2.1

Flags              :                    Type               : (*,G)

MRIB Next Hop      : 

MRIB Src Flags     : remote             Keepalive Timer    : Not Running

Up Time            : 0d 00:05:29        Resolved By        : unresolved

Up JP State        : Joined             Up JP Expiry       : 0d 00:00:30

Up JP Rpt          : Not Joined StarG   Up JP Rpt Override : 0d 00:00:00

Rpf Neighbor       : 

Incoming Intf      : mpls-if-73730

Outgoing Intf List : to-MC-client1

Curr Fwding Rate   : 14.7 kbps          

Forwarded Packets  : 12163              Discarded Packets  : 0

Forwarded Octets   : 559498             RPF Mismatches     : 0

Spt threshold      : 0 kbps             ECMP opt threshold : 7

Admin bandwidth    : 1 kbps             

-------------------------------------------------------------------------------

Groups : 1

===============================================================================

 

Influence IGP Metric

Suppose that the IGP metric is increased on all links pointing to/from PE-2 and on the link between PE-5 andPE-7.

A:PE-1# configure router ospf area 0 interface int-PE-1-PE-2 metric 10000

A:PE-2# configure router ospf area 0 interface int-PE-2-PE-4 metric 10000 

A:PE-2# configure router ospf area 0 interface int-PE-2-PE-3 metric 10000 

A:PE-2# configure router ospf area 0 interface int-PE-2-PE-1 metric 10000

A:PE-3# configure router ospf area 0 interface int-PE-3-PE-2 metric 10000

A:PE-4# configure router ospf area 0 interface int-PE-4-PE-2 metric 10000

A:PE-5# configure router ospf area 0 interface int-PE-5-PE-7 metric 10000

A:PE-7# configure router ospf area 0 interface int-PE-7-PE-5 metric 10000

The existing P2MP LSP LSP-p2mp-1 will not take into account these new constraints. The two S2L paths (one loose towards PE-6 and another one loose towards PE-7) are calculated using the default OSPF metric of 1000. What we can do to trigger MPLS to re-compute the S2L paths, is configure a p2mp-resignal-timer on the head-end node inside the global MPLS context. In this way, each time this timer expires (in our example, every 60 minutes), MPLS will trigger CSPF to re-compute the whole set of S2L paths of all active P2MP instance. MPLS performs a global make-before-break (MBB) and moves each S2L sub-LSP in the instance into its new path using a new P2MP LSP ID if the global MBB is successful. show router mpls status gives you an indication when the P2MP resignal timer will expire and which types of LSPs are setup on the node.

A:PE-1# configure router mpls p2mp-resignal-timer 60

A:PE-1# show router mpls status 

===============================================================================

MPLS Status

===============================================================================

Admin Status       : Up                 Oper Status        : Up

Oper Down Reason   : n/a                

FR Object          : Enabled            Resignal Timer     : Disabled

Hold Timer         : 1 seconds          Next Resignal      : N/A

Srlg Frr           : Disabled           Srlg Frr Strict    : Disabled

Dynamic Bypass     : Enabled            User Srlg Database : Disabled

Least Fill Min Thd.: 5 percent          LeastFill ReoptiThd: 10 percent

P2mp Resignal Timer: 60 minutes         P2mp Next Resignal : 6 minutes

Sec FastRetryTimer : Disabled           

LSP Counts          Originate           Transit             Terminate          

-------------------------------------------------------------------------------

…

===============================================================================

As an alternative the user can also perform a manual resignal of a P2MP instance on the head-end node using a tools command.

A:PE-1# tools perform router mpls resignal p2mp-lsp LSP-p2mp-1 p2mp-instance p-p-to-mp-1

A:PE-1# tools perform router mpls resignal p2mp-delay 0

 

Figure 171: P2MP LSP p-to-mp-1 with Metric Change

After an instantaneous tools resignal command with executed with no delay (p2mp-delay 0), the S2L paths can be verified will be according to Figure 171. PE-6 node is acting now as BUD LSR node (iso egress LER before)

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-1 p2mp-instance p-LSP-p2mp-1 s2l loose to 192.0.2.7 detail 

…

LSP Name    : LSP-p2mp-1                         S2l LSP ID  : 27658

P2MP ID     : 0                                  S2l Grp Id  : 2

Adm State   : Up                                 Oper State  : Up

S2l State:  : Active                                         :  

S2L Name    : loose                              To          : 192.0.2.7

S2l Admin   : Up                                 S2l Oper    : Up

OutInterface: 1/1/2                              Out Label   : 131065

S2L Up Time : 0d 02:14:31                        S2L Dn Time : 0d 00:00:00

RetryAttempt: 0                                  NextRetryIn : 0 sec

S2L Trans   : 6                                  CSPF Queries: 6

Failure Code: noError                            Failure Node: n/a

ExplicitHops:                                    

    No Hops Specified

Actual Hops :                                    

    192.168.13.1(192.0.2.1) @                    Record Label    : N/A

 -> 192.168.13.2(192.0.2.3) @                    Record Label    : 131065

 -> 192.168.35.2(192.0.2.5) @                    Record Label    : 131062

 -> 192.168.45.1(192.0.2.4) @                    Record Label    : 131061

 -> 192.168.46.2(192.0.2.6) @                    Record Label    : 131065

 -> 192.168.67.2(192.0.2.7)                      Record Label    : 131063

ComputedHops:                                    

    192.168.13.1    -> 192.168.13.2    -> 192.168.35.2    -> 192.168.45.1   

 -> 192.168.46.2    -> 192.168.67.2   

LastResignal: n/a                                

==================================================================

A:PE-1# 

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-1 p2mp-instance p-LSP-p2mp-1 s2l loose to 192.0.2.6 detail 

…

LSP Name    : LSP-p2mp-1                         S2l LSP ID  : 27658

P2MP ID     : 0                                  S2l Grp Id  : 1

Adm State   : Up                                 Oper State  : Up

S2l State:  : Active                                         :  

S2L Name    : loose                              To          : 192.0.2.6

S2l Admin   : Up                                 S2l Oper    : Up

OutInterface: 1/1/2                              Out Label   : 131065

S2L Up Time : 0d 02:22:00                        S2L Dn Time : 0d 00:00:00

RetryAttempt: 0                                  NextRetryIn : 0 sec

S2L Trans   : 6                                  CSPF Queries: 6

Failure Code: noError                            Failure Node: n/a

ExplicitHops:                                    

    No Hops Specified

Actual Hops :                                    

    192.168.13.1(192.0.2.1) @                    Record Label    : N/A

 -> 192.168.13.2(192.0.2.3) @                    Record Label    : 131065

 -> 192.168.35.2(192.0.2.5) @                    Record Label    : 131062

 -> 192.168.45.1(192.0.2.4) @                    Record Label    : 131061

 -> 192.168.46.2(192.0.2.6)                      Record Label    : 131065

ComputedHops:                                    

    192.168.13.1    -> 192.168.13.2    -> 192.168.35.2    -> 192.168.45.1   

 -> 192.168.46.2   

LastResignal: n/a                                

==================================================================

A:PE-1# 

 

An oam p2mp-lsp-trace command toward PE-7 will now set the E flag on PE-6 since that PE acts also as an egress LER node.

A:PE-1# oam p2mp-lsp-trace LSP-p2mp-1 p2mp-instance p-LSP-p2mp-1 s2l-dest-address 192.0.2.7 detail 

P2MP LSP LSP-p2mp-1: 128 bytes MPLS payload

P2MP Instance p-LSP-p2mp-1, S2L Egress 192.0.2.7

 1  192.0.2.3  rtt=3.08 ms rc=8(DSRtrMatchLabel)

     DS 1: IfAddr 192.168.35.2 MRU=1500 label=131062 proto=4(RSVP-TE) B/E flags:0/0 

  2  192.0.2.5  rtt=3.56 ms rc=8(DSRtrMatchLabel)

     DS 1: IfAddr 192.168.45.1 MRU=1500 label=131061 proto=4(RSVP-TE) B/E flags:0/0 

  3  192.0.2.4  rtt=11.6 ms rc=8(DSRtrMatchLabel)

     DS 1: IfAddr 192.168.46.2 MRU=1500 label=131065 proto=4(RSVP-TE) B/E flags:0/0 

  4  192.0.2.6  rtt=4.84 ms rc=8(DSRtrMatchLabel)

     DS 1: IfAddr 192.168.67.2 MRU=1500 label=131063 proto=4(RSVP-TE) B/E flags:0/1 

  5  192.0.2.7  rtt=6.05 ms rc=3(EgressRtr)

As a next step, the S2L path towards PE-7 is changed from loose to a strict direct MPLS path (strict-to-PE-7). In that way, OSPF is not calculating anymore the shortest path to the leaf node.

A:PE-1# configure router mpls 

            …

            path "strict-to-PE-7"

                hop 10 192.168.13.2 strict

                hop 20 192.168.35.2 strict

                hop 30 192.168.57.2 strict

                no shutdown

            exit

Before applying this new S2L path to the existing P2MP LSP (LSP-p2mp-1), the existing S2L path towards PE-7 must be removed.

A:PE-1# configure router mpls lsp "LSP-p2mp-1" primary-p2mp-instance "p-LSP-p2mp-1" s2l-path "loose" to 192.0.2.7 shutdown

A:PE-1# configure router mpls lsp "LSP-p2mp-1" primary-p2mp-instance "p-LSP-p2mp-1" no s2l-path "loose" to 192.0.2.7

A:PE-1# configure router mpls lsp "LSP-p2mp-1" primary-p2mp-instance "p-LSP-p2mp-1" s2l-path "strict-to-PE-7" to 192.0.2.7  

Figure 172: P2MP LSP LSP-p2mp-1 with Strict S2L Path Towards PE-7

As a consequence of this, only the S2l Grp Id has changed while S2L LSP ID remains the same as before. Now, S2L paths can be verified according to Figure 172. PE-5 is acting now as branch LSR node (iso transit LSR before).

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-1 p2mp-instance p-LSP-p2mp-1 s2l strict-to-PE-7 to 192.0.2.7 detail 

===============================================================================

MPLS LSP LSP-p2mp-1 S2L strict-to-PE-7 (Detail)

===============================================================================

Legend : 

    @ - Detour Available                          # - Detour In Use

    b - Bandwidth Protected                       n - Node Protected

    s - Soft Preemption                   

===============================================================================

-------------------------------------------------------------------------------

LSP LSP-p2mp-1 S2L strict-to-PE-7

-------------------------------------------------------------------------------

LSP Name    : LSP-p2mp-1                         S2l LSP ID  : 27658

P2MP ID     : 0                                  S2l Grp Id  : 3

Adm State   : Up                                 Oper State  : Up

S2l State:  : Active                                         :  

S2L Name    : strict-to-PE-7                     To          : 192.0.2.7

S2l Admin   : Up                                 S2l Oper    : Up

OutInterface: 1/1/2                              Out Label   : 131065

S2L Up Time : 0d 00:01:04                        S2L Dn Time : 0d 00:00:00

RetryAttempt: 0                                  NextRetryIn : 0 sec

S2L Trans   : 1                                  CSPF Queries: 1

Failure Code: noError                            Failure Node: n/a

ExplicitHops:                                    

    192.168.13.2    -> 192.168.35.2    -> 192.168.57.2   

Actual Hops :                                    

    192.168.13.1(192.0.2.1) @                    Record Label    : N/A

 -> 192.168.13.2(192.0.2.3) @                    Record Label    : 131065

 -> 192.168.35.2(192.0.2.5) @                    Record Label    : 131062

 -> 192.168.57.2(192.0.2.7)                      Record Label    : 131068

ComputedHops:                                    

    192.168.13.1    -> 192.168.13.2    -> 192.168.35.2    -> 192.168.57.2   

LastResignal: n/a                                

===============================================================================

An oam p2mp-lsp-trace command towards PE-7 will now set the B flag on PE-5 since that became a branch LSR now.

A:PE-1# oam p2mp-lsp-trace LSP-p2mp-1 p2mp-instance p-LSP-p2mp-1 s2l-dest-address 192.0.2.7 detail 

P2MP LSP LSP-p2mp-1: 128 bytes MPLS payload

P2MP Instance p-LSP-p2mp-1, S2L Egress 192.0.2.7

  1  192.0.2.3  rtt=2.67 ms rc=8(DSRtrMatchLabel)

     DS 1: IfAddr 192.168.35.2 MRU=1500 label=131062 proto=4(RSVP-TE) B/E flags:0/0 

  2  192.0.2.5  rtt=3.35 ms rc=8(DSRtrMatchLabel)

     DS 1: IfAddr 192.168.57.2 MRU=1500 label=131068 proto=4(RSVP-TE) B/E flags:1/0 

  3  192.0.2.7  rtt=5.35 ms rc=3(EgressRtr)

 

Intelligent Remerge

Intelligent remerge protects users from receiving duplicate multicast during convergence. It also protects against duplicate traffic in case of badly designed S2L paths. Initially, three cases exist for which intelligent remerge is implemented.

Case 1

When the paths of two different S2Ls of the same P2MP LSP instance have Ingress Label Maps (ILMs) on different ports but go out on the same Next-hop Label Forwarding Entry (NHLFE).

Figure 173: Intelligent Remerge, Case 1

On the head-end node (PE-1), a new P2MP LSP (‘LSP-p2mp-2’) will be created with two strict direct MPLS paths (‘strict-to-PE-7’ and ‘strict-to-PE-6’). See Figure 173 for detailed address information. Intelligent re-merge is performed at node PE-5.

 

 

A:PE-1# configure router mpls 

            …

            path "strict-to-PE-7"

                hop 10 192.168.13.2 strict

                hop 20 192.168.35.2 strict

                hop 30 192.168.57.2 strict

                no shutdown

            exit

            path "strict-to-PE-6"

                hop 10 192.168.12.2 strict

                hop 20 192.168.24.2 strict

                hop 30 192.168.45.2 strict

                hop 40 192.168.57.2 strict

                hop 50 192.168.67.1 strict

                no shutdown

            exit

            …

            lsp "LSP-p2mp-2" p2mp-lsp

                primary-p2mp-instance "p-LSP-p2mp-2"

                    s2l-path "strict-to-PE-7" to 192.0.2.7

                    exit

                    s2l-path "strict-to-PE-6" to 192.0.2.6

                    exit

                exit

                no shutdown

            exit

            no shutdown

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-2 p2mp-instance p-LSP-p2mp-2 

===============================================================================

MPLS P2MP Instance (Originating)

===============================================================================

-------------------------------------------------------------------------------

Type : Originating 

-------------------------------------------------------------------------------

LSP Name    : LSP-p2mp-2                       LSP Tunnel ID  : 13

P2MP ID     : 0                                

Adm State   : Up                               Oper State     : Up

P2MPInstance: p-LSP-p2mp-2                     P2MP-Inst-type : Primary

P2MP Inst Id: 4                                P2MP Lsp Id    : 22016

Inst Admin  : Up                               Inst Oper      : Up

Inst Up Time: 0d 00:01:56                      Inst Dn Time   : 0d 00:00:00

Hop Limit   : 255                              Adaptive       : Enabled

Record Route: Record                           Record Label   : Record

Include Grps:                                  Exclude Grps   :  

None                                           None

Bandwidth   : No Reservation                   Oper Bw        : 0 Mbps

S2l-Name    : strict-to-PE-7                   To             : 192.0.2.7

S2l Admin   : Up                               S2l Oper       : Up

S2l-Name    : strict-to-PE-6                   To             : 192.0.2.6

S2l Admin   : Up                               S2l Oper       : Up

-------------------------------------------------------------------------------

P2MP instances : 1

===============================================================================

 

To verify that PE-5 node is not sending duplicate multicast traffic downstream towards PE-7 while it receives two incoming multicast streams, a new tunnel interface and a new static <S,G> IGMP join will be configured on head-end node (PE-1) and leaf nodes (PE-6 and PE-7). Also on the leaf nodes, an extension to the existing multicast information policy is needed. Translated into configuration commands, this becomes

A:PE-1# configure router pim tunnel-interface rsvp-p2mp LSP-p2mp-2

A:PE-1# configure router igmp 

        …

        tunnel-interface rsvp-p2mp "LSP-p2mp-2" 

            static

                group 227.2.2.2

                    source 192.168.9.2

A:PE-6# configure router pim tunnel-interface rsvp-p2mp LSP-p2mp-2 sender 192.0.2.1

A:PE-6# configure router igmp interface to-MC-client2 static group 227.2.2.2 source 192.168.9.2

A:PE-6# configure mcast-management 

            multicast-info-policy "p2mp-pol" create

                bundle "bundle2" create

                    primary-tunnel-interface rsvp-p2mp LSP-p2mp-2 sender 192.0.2.1

                    channel "227.2.2.2" "227.2.2.2" create

                    exit

                exit

A:PE-7# configure router pim tunnel-interface rsvp-p2mp LSP-p2mp-2 sender 192.0.2.1

A:PE-7# configure router igmp interface to-MC-client1 static group 227.2.2.2 source 192.168.9.2

A:PE-7# configure mcast-management 

            multicast-info-policy "p2mp-pol" create

                bundle "bundle2" create

                    primary-tunnel-interface rsvp-p2mp LSP-p2mp-2 sender 192.0.2.1

                    channel "227.2.2.2" "227.2.2.2" create

                    exit

                exit

 

For verification of incoming/outgoing multicast traffic at PE-5 node, the monitor command is used.

A:PE-5# monitor port 1/1/1 1/1/2 1/1/3 rate interval 3 repeat 100 

===============================================================================

Monitor statistics for Ports

===============================================================================

                                                   Input                 Output

-------------------------------------------------------------------------------

…

-------------------------------------------------------------------------------

At time t = 12 sec (Mode: Rate)

-------------------------------------------------------------------------------

Port 1/1/1

-------------------------------------------------------------------------------

Octets                                              2714                     50

Packets                                               40                      0

Errors                                                 0                      0

Utilization (% of port capacity)                    0.02                  ~0.00

Port 1/1/2

-------------------------------------------------------------------------------

Octets                                              2769                     50

Packets                                               40                      0

Errors                                                 0                      0

Utilization (% of port capacity)                    0.02                  ~0.00

Port 1/1/3

-------------------------------------------------------------------------------

Octets                                                21                   2714

Packets                                                0                     40

Errors                                                 0                      0

Utilization (% of port capacity)                   ~0.00                   0.02

As a conclusion we can say that two incoming multicast streams are seen at PE-5 node (port 1/1/2 and port 1/1/1) and only one outgoing multicast stream (port 1/1/3) is sent. No traffic duplication is seen.

 

Case 2

When two paths of the same S2L have ILMs on different incoming ports and go out on the same NHLFE. This is the case when we perform make-before-break (MBB) on an S2L path due to graceful shutdown or global revertive. This is only a temporary situation since the original path will be torn down.

Figure 174: Intelligent Remerge, Case 2

For this test, only one MC-client will be looked at (the one connected to head-end node PE-7). On PE-4 and PE-7 nodes, port 1/1/4 will be shutdown to isolate PE-6. On the head-end node (PE-1), a new P2MP LSP (LSP-p2mp-3) will be created with one loose MPLS path (loose). See Figure 173 for detailed address information. Also in this case, intelligent re-merge is performed at node PE-5.

A:PE-1# configure router mpls 

            …

            path "loose"

                no shutdown

            exit

            …

            lsp "LSP-p2mp-3" p2mp-lsp

                primary-p2mp-instance "p-LSP-p2mp-3"

                    s2l-path "loose" to 192.0.2.7

                    exit

                exit

                no shutdown

            exit

            no shutdown

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-3 p2mp-instance p-LSP-p2mp-3 s2l loose to 192.0.2.7 detail 

===============================================================================

MPLS LSP LSP-p2mp-3 S2L loose (Detail)

===============================================================================

Legend : 

    @ - Detour Available                          # - Detour In Use

    b - Bandwidth Protected                       n - Node Protected

    s - Soft Preemption                   

===============================================================================

-------------------------------------------------------------------------------

LSP LSP-p2mp-3 S2L loose

-------------------------------------------------------------------------------

LSP Name    : LSP-p2mp-3                         S2l LSP ID  : 17410

P2MP ID     : 0                                  S2l Grp Id  : 5

Adm State   : Up                                 Oper State  : Up

S2l State:  : Active                                         :  

S2L Name    : loose                              To          : 192.0.2.7

S2l Admin   : Up                                 S2l Oper    : Up

OutInterface: 1/1/2                              Out Label   : 131069

S2L Up Time : 0d 00:00:27                        S2L Dn Time : 0d 00:00:00

RetryAttempt: 0                                  NextRetryIn : 0 sec

S2L Trans   : 3                                  CSPF Queries: 0

Failure Code: noError                            Failure Node: n/a

ExplicitHops:                                    

    No Hops Specified

Actual Hops :                                    

    192.168.13.1(192.0.2.1)                      Record Label    : N/A

 -> 192.168.13.2(192.0.2.3)                      Record Label    : 131069

 -> 192.168.35.2(192.0.2.5)                      Record Label    : 131068

 -> 192.168.57.2(192.0.2.7)                      Record Label    : 131071

LastResignal: n/a                                

===============================================================================

 

In a normal situation, the P2MP LSP would follow the nodes PE-1, PE-3, PE-5 and PE-7. This can be verified with MC traffic. Therefore, a new tunnel interface and a new static <S,G> IGMP join will be configured on head-end node (PE-1) and leaf node (PE-7). On the leaf node, an extension to the existing multicast information policy is needed. Translated into configuration commands, this becomes

 A:PE-1# configure router pim tunnel-interface rsvp-p2mp LSP-p2mp-3

A:PE-1# configure router igmp 

        …

        tunnel-interface rsvp-p2mp "LSP-p2mp-3" 

            static

                group 227.3.3.3

                    source 192.168.9.2

A:PE-7# configure router pim tunnel-interface rsvp-p2mp LSP-p2mp-3 sender 192.0.2.1

A:PE-7# configure router igmp interface to-MC-client1 static group 227.3.3.3 source 192.168.9.2

A:PE-7# configure mcast-management 

            multicast-info-policy "p2mp-pol" create

                bundle "bundle3" create

                    primary-tunnel-interface rsvp-p2mp LSP-p2mp-3 sender 192.0.2.1

                    channel "227.3.3.3" "227.3.3.3" create

                    exit

                exit

A:PE-5# monitor port 1/1/1 1/1/2 1/1/3 rate interval 3 repeat 999

===============================================================================

Monitor statistics for Ports

===============================================================================

                                                   Input                 Output

-------------------------------------------------------------------------------

…

-------------------------------------------------------------------------------

At time t = 231 sec (Mode: Rate)

-------------------------------------------------------------------------------

Port 1/1/1

-------------------------------------------------------------------------------

Octets                                              2714                    108

Packets                                               40                      1

Errors                                                 0                      0

Utilization (% of port capacity)                    0.02                  ~0.00

-------------------------------------------------------------------------------

Port 1/1/2

-------------------------------------------------------------------------------

Octets                                                50                     50

Packets                                                0                      0

Errors                                                 0                      0

Utilization (% of port capacity)                   ~0.00                  ~0.00

Port 1/1/3

-------------------------------------------------------------------------------

Octets                                                81                   2801

Packets                                                1                     40

Errors                                                 0                      0

Utilization (% of port capacity)                   ~0.00                   0.02

Now, perform an RSVP graceful shut on PE-3 node. Global revertive is triggered on head-end node (PE-1). A new MPLS path will be calculated (see dotted line in Figure 173). For a few seconds, the old path and new path are active (two incoming MC streams on PE-5 node). PE-5 node is doing intelligent re-merge, not sending duplicate multicast traffic downstream towards PE-7 :

A:PE-3# configure router rsvp graceful-shutdown

A:PE-5# monitor port 1/1/1 1/1/2 1/1/3 rate interval 3 repeat 999

===============================================================================

Monitor statistics for Ports

===============================================================================

                                                   Input                 Output

-------------------------------------------------------------------------------

…

-------------------------------------------------------------------------------

At time t = 31 sec (Mode: Rate)

-------------------------------------------------------------------------------

Port 1/1/1

-------------------------------------------------------------------------------

Octets                                              2714                    108

Packets                                               40                      1

Errors                                                 0                      0

Utilization (% of port capacity)                    0.02                  ~0.00

-------------------------------------------------------------------------------

Port 1/1/2

-------------------------------------------------------------------------------

Octets                                              2714                    108

Packets                                               40                      1

Errors                                                 0                      0

Utilization (% of port capacity)                   ~0.00                  ~0.00

Port 1/1/3

-------------------------------------------------------------------------------

Octets                                                81                   2801

Packets                                                1                     40

Errors                                                 0                      0

Utilization (% of port capacity)                   ~0.00                   0.02

No traffic duplication is seen.

 

Case 3

When a bypass is active on the S2L path and the new global revertive path of the same S2L arrives on the same incoming interface as the original path (interface flapped) at the FRR merge point node. The implementation recognizes this specific case and will signal a different label from the original S2L path coming on that same interface.

 

Figure 175: Intelligent Remerge, Case 3

For this test, all the non-default OSPF metrics are removed from the interfaces. Only one MC-client will be looked at (the one connected to head-end node PE-7). On PE-4 and PE-7 node, port 1/1/4 will be shutdown to isolate PE-6. On the head-end node (PE-1), a new P2MP LSP (LSP-p2mp-4) will be created with one loose MPLS path (loose) and FRR enabled. See Figure 175 for detailed address information. Also in this case, intelligent re-merge is performed at node PE-5.

A:PE-1# configure router mpls 

            …

            path "loose"

                no shutdown

            exit

            …

            lsp "LSP-p2mp-4" p2mp-lsp

                cspf

                fast-reroute facility

                    no node-protect

                exit

                primary-p2mp-instance "p-LSP-p2mp-4"

                    s2l-path "loose" to 192.0.2.7

                    exit

                exit

                no shutdown

            exit

            no shutdown

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-4 p2mp-instance p-LSP-p2mp-4 s2l loose to 192.0.2.7 detail 

===============================================================================

MPLS LSP LSP-p2mp-4 S2L loose (Detail)

===============================================================================

Legend : 

    @ - Detour Available                          # - Detour In Use

    b - Bandwidth Protected                       n - Node Protected

    s - Soft Preemption                   

===============================================================================

-------------------------------------------------------------------------------

LSP LSP-p2mp-4 S2L loose

-------------------------------------------------------------------------------

LSP Name    : LSP-p2mp-4                         S2l LSP ID  : 39940

P2MP ID     : 0                                  S2l Grp Id  : 4

Adm State   : Up                                 Oper State  : Up

S2l State:  : Active                                         :  

S2L Name    : loose                              To          : 192.0.2.7

S2l Admin   : Up                                 S2l Oper    : Up

OutInterface: 1/1/2                              Out Label   : 131069

S2L Up Time : 0d 00:00:31                        S2L Dn Time : 0d 00:00:00

RetryAttempt: 0                                  NextRetryIn : 0 sec

S2L Trans   : 1                                  CSPF Queries: 1

Failure Code: noError                            Failure Node: n/a

ExplicitHops:                                    

    No Hops Specified

Actual Hops :                                    

    192.168.13.1(192.0.2.1) @                    Record Label    : N/A

 -> 192.168.13.2(192.0.2.3) @                    Record Label    : 131069

 -> 192.168.35.2(192.0.2.5)                      Record Label    : 131069

 -> 192.168.57.2(192.0.2.7)                      Record Label    : 131070

ComputedHops:                                    

    192.168.13.1    -> 192.168.13.2    -> 192.168.35.2    -> 192.168.57.2   

LastResignal: n/a                                

===============================================================================

 

In the normal situation, the P2MP LSP follows the nodes PE-1, PE-3, PE-5 and PE-7. This can be verified with MC traffic. Therefore, a new tunnel interface and a new static <S,G> IGMP join will be configured on head-end node (PE-1) and leaf node (PE-7). On the leaf node, an extension to the existing multicast information policy is needed. Translated into configuration commands, this becomes:

A:PE-1# configure router pim tunnel-interface rsvp-p2mp LSP-p2mp-4

A:PE-1# configure router igmp 

        …

        tunnel-interface rsvp-p2mp "LSP-p2mp-4" 

            static

                group 227.4.4.4

                    source 192.168.9.2

A:PE-7# configure router pim tunnel-interface rsvp-p2mp LSP-p2mp-4 sender 192.0.2.1

A:PE-7# configure router igmp interface to-MC-client1 static group 227.4.4.4 source 192.168.9.2

A:PE-7# configure mcast-management 

            multicast-info-policy "p2mp-pol" create

                bundle "bundle4" create

                    primary-tunnel-interface rsvp-p2mp LSP-p2mp-4 sender 192.0.2.1

                    channel "227.4.4.4" "227.4.4.4" create

                    exit

                exit

A:PE-5# monitor port 1/1/1 1/1/2 1/1/3 rate interval 3 repeat 999

===============================================================================

Monitor statistics for Ports

===============================================================================

                                                   Input                 Output

-------------------------------------------------------------------------------

…

-------------------------------------------------------------------------------

At time t = 3 sec (Mode: Rate)

-------------------------------------------------------------------------------

Port 1/1/1

-------------------------------------------------------------------------------

Octets                                              2737                     21

Packets                                               40                      0

Errors                                                 0                      0

Utilization (% of port capacity)                    0.02                  ~0.00

Port 1/1/2

-------------------------------------------------------------------------------

Octets                                                21                     21

Packets                                                0                      0

Errors                                                 0                      0

Utilization (% of port capacity)                   ~0.00                  ~0.00

Port 1/1/3

-------------------------------------------------------------------------------

Octets                                                21                   2714

Packets                                                0                     40

Errors                                                 0                      0

Utilization (% of port capacity)                   ~0.00                   0.02

Now we do an link failure on PE-3 <=> PE-5 interface.

A:PE-3# configure port 1/1/1 shutdown

As a consequence of this, traffic will be flowing over the bypass link (see also Figure 175 + ‘#’ symbol in next show command).

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-4 p2mp-instance p-LSP-p2mp-4 s2l loose to 192.0.2.7 detail 

===============================================================================

MPLS LSP LSP-p2mp-4 S2L loose (Detail)

===============================================================================

Legend : 

    @ - Detour Available                          # - Detour In Use

    b - Bandwidth Protected                       n - Node Protected

    s - Soft Preemption                   

===============================================================================

-------------------------------------------------------------------------------

LSP LSP-p2mp-4 S2L loose

-------------------------------------------------------------------------------

LSP Name    : LSP-p2mp-4                         S2l LSP ID  : 39940

P2MP ID     : 0                                  S2l Grp Id  : 4

Adm State   : Up                                 Oper State  : Up

S2l State:  : Active                                         :  

S2L Name    : loose                              To          : 192.0.2.7

S2l Admin   : Up                                 S2l Oper    : Up

OutInterface: 1/1/2                              Out Label   : 131069

S2L Up Time : 0d 00:04:21                        S2L Dn Time : 0d 00:00:00

RetryAttempt: 0                                  NextRetryIn : 0 sec

S2L Trans   : 1                                  CSPF Queries: 1

Failure Code: tunnelLocallyRepaired              Failure Node: 192.0.2.3

ExplicitHops:                                    

    No Hops Specified

Actual Hops :                                    

    192.168.13.1(192.0.2.1) @                    Record Label    : N/A

 -> 192.168.13.2(192.0.2.3) @ #                  Record Label    : 131069

 -> 192.168.35.2(192.0.2.5)                      Record Label    : 131069

 -> 192.168.57.2(192.0.2.7)                      Record Label    : 131070

ComputedHops:                                    

    192.168.13.1    -> 192.168.13.2    -> 192.168.35.2    -> 192.168.57.2   

LastResignal: n/a                                

In Prog MBB :

 MBB Type   : GlobalRevert                       NextRetryIn : 1 sec

 Started At : 11/13/2009 14:46:50                RetryAttempt: 0

 FailureCode: noError                            Failure Node: n/a

===============================================================================

In the meantime, PE-3 will trigger a global revertive action (sending PATHerr message) towards the head-end node (PE-1).

A:PE-1# show router mpls p2mp-lsp LSP-p2mp-4 p2mp-instance p-LSP-p2mp-4 s2l loose to 192.0.2.7 detail 

===============================================================================

MPLS LSP LSP-p2mp-4 S2L loose (Detail)

===============================================================================

Legend : 

    @ - Detour Available                          # - Detour In Use

    b - Bandwidth Protected                       n - Node Protected

    s - Soft Preemption                   

===============================================================================

-------------------------------------------------------------------------------

LSP LSP-p2mp-4 S2L loose

-------------------------------------------------------------------------------

LSP Name    : LSP-p2mp-4                         S2l LSP ID  : 39940

P2MP ID     : 0                                  S2l Grp Id  : 5

Adm State   : Up                                 Oper State  : Up

S2l State:  : Active                                         :  

S2L Name    : loose                              To          : 192.0.2.7

S2l Admin   : Up                                 S2l Oper    : Up

OutInterface: 1/1/3                              Out Label   : 131071

S2L Up Time : 0d 00:06:44                        S2L Dn Time : 0d 00:00:00

RetryAttempt: 0                                  NextRetryIn : 0 sec

S2L Trans   : 2                                  CSPF Queries: 2

Failure Code: noError                            Failure Node: n/a

ExplicitHops:                                    

    No Hops Specified

Actual Hops :                                    

    192.168.12.1(192.0.2.1) @                    Record Label    : N/A

 -> 192.168.12.2(192.0.2.2)                      Record Label    : 131071

 -> 192.168.24.2(192.0.2.4)                      Record Label    : 131071

 -> 192.168.45.2(192.0.2.5)                      Record Label    : 131071

 -> 192.168.57.2(192.0.2.7)                      Record Label    : 131070

ComputedHops:                                    

    192.168.12.1    -> 192.168.12.2    -> 192.168.24.2    -> 192.168.45.2   

 -> 192.168.57.2   

LastResignal: n/a                                

Last MBB    :

 MBB Type   : GlobalRevert                       MBB State   : Success

 Ended At   : 11/13/2009 14:47:23                

===============================================================================

 

For some time, PE-5 will receive two incoming MC streams (both arriving on port 1/1/2). One from ‘bypass’ path (PE-3 => PE-2 => PE-4 => PE-5) and one from new MPLS path ( PE-1 => PE-2 => PE-4 => PE-5 => PE-7). Port 1/1/3 on PE-5 performs intelligent remerge, only one MC stream is sent downstream towards leaf node PE-7.

A:PE-5# monitor port 1/1/1 1/1/2 1/1/3 rate interval 3 repeat 999

===============================================================================

Monitor statistics for Ports

===============================================================================

                                                   Input                 Output

-------------------------------------------------------------------------------

…

-------------------------------------------------------------------------------

At time t = 39 sec (Mode: Rate)

-------------------------------------------------------------------------------

Port 1/1/1

-------------------------------------------------------------------------------

Octets                                                 0                    261

Packets                                                0                      1

Errors                                                 0                      0

Utilization (% of port capacity)                    0.00                  ~0.00

Port 1/1/2

-------------------------------------------------------------------------------

Octets                                              5540                    393

Packets                                               80                      2

Errors                                                 0                      0

Utilization (% of port capacity)                    0.04                  ~0.00

Port 1/1/3

-------------------------------------------------------------------------------

Octets                                               128                   3041

Packets                                                1                     40

Errors                                                 0                      0

Utilization (% of port capacity)                   ~0.00                   0.02

-------------------------------------------------------------------------------

Conclusion

Looking to the configuration point of view, a P2MP LSP is only configured on the head-end node of that P2MP LSP, no explicit configuration is needed on the transit LSRs, branch LSRs, BUD LSRs and egress LERs/leaf nodes.

Since PIM protocol is only needed on the head-end node and leaf node(s), we can work in a PIM-free core network. Although convergence is not covered in this configuration note, failures in the core will be resolved by MPLS (in case of FRR, traffic loss < 50ms is expected). This is a major improvement compared to PIM convergence.

 

 

---

1

Convergence is out of the scope of this configuration note.

2

As long as one S2L path is operationally up (show router mpls p2mp-lsp lsp-name p2mp-instance instance-name) , the Oper State of the P2MP LSP is Up.