7-7 IP Multicast Routing | Cisco Field Manual[c] Router Configuration

IP multicast routing provides multicast routing support using PIM (Protocol-Independent Multicasting) Version 1 or Version 2.
It operates in dense mode, sparse mode, or a combination of both.
It supports automatic rendezvous points (RPs) for sparse-mode PIM.
It uses IGMP to communicate with the multicast host for the reporting of group memberships.
It supports IGMP Version 1 and Version 2.
It provides Cisco Group Management Protocol (CGMP) support to provide Cisco switches with Layer 2 to Layer 3 multicast information.

NOTE

IGMP is defined by RFC 1112 and provides a mechanism through which multicast servers can advertise a group address and clients can subscribe to that group. These groups are learned by the multicast routing protocol and are forwarded between participating routers to provide pathways for the groups. IGMP uses the multicast addresses 224.0.0.1 for all systems on a subnet and 224.0.0.2 for all routers on the subnet.

Configuration

Enable IP multicast routing:
```
 (global)  ip multicast-routing  
```
This command enables the multicast routing capabilities on the router. If you do not enable multicast routing, none of the PIM commands have any effect.
Configure PIM on a participating multicast interface:
```
 (interface)  ip pim  {  dense-mode   sparse-mode   sparse-dense-mode  } 
```
Any interface that will receive or forward multicast traffic on the connected network must be configured with PIM. Which mode you choose for that network depends on the traffic requirements. Dense mode operation requires that PIM be configured on each segment that will pass the multicast traffic. Sparse mode operation allows you to run multicast packets on only the segments that will have multicast servers. Which mode you choose to configure depends on the particulars of your network. The command option sparse-dense-mode is the most versatile. With this option, the interface operates in dense mode for some groups on the connected network and sparse mode for other groups. In order to have any groups operate in sparse mode, you must configure a rendezvous point.
(Required for sparse mode) Configure a rendezvous point:
```
 (global)  ip pim rp-address   ip-address  [  access-list-number  ] [  override  ] 
```
This command specifies the RP for leaf routers so that they can register groups with the RP or determine the location of group routers from the RP. The access-list-number parameter allows you to configure specific groups for a given RP. By using the command with an access list specifying the group address, you can configure multiple RPsone for each access list configured. The override option tells the router to use the configured group over one that is learned with the auto RP option (discussed in the next step).
(Optional) Configure automatic rendezvous points.
1. Configure a router to be an automatic RP:
```
 (global)  ip pim send-rp-announce   type number   scope   ttl   group-list   access-list-number  
```
  The router announces itself as an RP to the mapping agent for the groups defined by the group-list access-list-number option. The type and number options identify the interface address that is announced as the RP address; this must be a PIM interface. The scope option determines how far in hops the message will be propagated.
2. Configure the RP mapping agent:
```
 (global)  ip pim send-rp-discovery scope   ttl  
```
  This configures a router that sends discovery packets telling other routes which group-to-RP mappings to use. The scope option determines how far in hops the message will be propagated (specified by the ttl parameter).
3. Filter incoming RP messages:
```
 (global)  ip pim rp-announce-filter rp-list   access-list-number   group-list   access-list-number  
```
  The rp-list option is used to filter incoming messages from a particular RP. The group-list option filters messages about the listed group.
(Optional) Configure the PIM version:
```
 (interface)  ip pim version  [  1   2  ] 
```
Specify the version of PIM operating on an interface. This command is used to reenable PIM Version 2 or specify PIM Version 1.

NOTE

Version 2 is the default for routers using Cisco IOS Software Release 11.3(2)T or later. All routers in the same broadcast domain must be the same version of PIM. A PIM Version 2 router automatically downgrades itself if it detects PIM Version 1 messages.
(Required for PIM Version 2) Configure candidate BSRs:
```
 (global)  ip pim bsr-candidate   type number hash-mask-length  [  priority  ] 
```
PIM Version 2 is an IETF standards track protocol. It uses a mechanism called a bootstrap router (BSR) to discover and announce the RP-set information. This command configures candidates to become BSRs. The type and number specify the PIM interface from which the BSR address is derived. The hash-mask-length is a value up to 32 bits that is ANDed with the group address to allow you to specify which parts of the address matter. This is used to get one RP for multiple groups. The optional priority setting allows you to increase the priority to determine which router will be chosen as the BSR. The priority is an integer from 0 to 255. The bootstrap router with the larger priority is preferred. If the priority values are the same, the router with the larger IP address is the bootstrap router. The default value is 0.
(Required for Version 2) Configure candidate RPs:
```
 (global)  ip pim rp-candidate   type number ttl   group-list   access-list-number  
```
This specifies a router to be an RP for all or any portion of the multicast groups. The type and number specify which PIM interface address will be used to identify the RP. ttl describes how many hops the messages can be propagated. The group-list option allows you to specify which multicast groups this RP serves.
(Optional) Configure a PIM domain border:
```
 (interface)  ip pim border  
```
This command is placed on an interface to set a border for the PIM domain. Bootstrap messages will not cross this interface in either direction, allowing for the configuration of different BSRs on either side of the border.
(Optional) Define the IP multicast boundary:
```
 (global)  ip multicast boundary   access-list-number  
```
Use this command along with an access list specifying groups 224.0.1.39 and 224.0.1.40 to prevent auto RP messages from entering into a PIM Version 2 domain with a BSR.
(Optional) Configure a router to be a member of a multicast group:
```
 (interface)  ip igmp join-group   group-address  
```
This command allows you to specify a multicast group for an interface to join. By joining a group, a router interface becomes a pingable member of the multicast group. This is a valuable testing tool.
(Optional) Control the groups that hosts on an interface can join:
```
 (interface)  ip igmp access-group   access-list-number  
```
This command allows you to control which groups the router presents to the clients for membership. By placing restrictions on groups, the interface does not allow the clients to join the groups denied by the access list.
(Optional) Configure a router as a statically connected member:
```
 (interface)  ip igmp static-group   group-address  
```
This command, like the ip igmp join-group command, allows an interface to become a member of the group. The difference is that the static group forwards only packets belonging to a group and does not accept (respond to) these packets. This allows the group packets to be fast-switched, because the interface doesn't have to accept the packets.
(Optional) Configure IGMP.

IGMP is used to communicate with the group devices. Because each segment might not have any clients, IGMP does not have to be configured. IGMP does run by default. On segments where clients are present, it might be important to configure some of the following options:
1. Specify the IGMP version:
```
 (interface)  ip igmp version  [  2   1  ] 
```
  This command specifies which version of IGMP the interface runs. The default is Version 2, but you should configure the router for Version 1 if your hosts do not support Version 2.
2. Set the IGMP query interval:
```
 (interface)  ip igmp query-interval   seconds  
```
  This specifies how often, in seconds, the IGMP designated router sends query messages to the hosts to determine group memberships. The default is 60 seconds.
3. Set the IP IGMP query timeout:
```
 (interface)  ip igmp query-timeout   seconds  
```
  This specifies how long the router waits before it takes over as the querier for a particular network. By default, it is two times the query interval, but this command allows you to change the timeout to the number of seconds specified.
4. Set the maximum query response time:
```
 (interface)  ip igmp query-max-response-time   seconds  
```
  This command is used to determine how long the router waits for a response from a host about group membership. The default is 10 seconds, but you can alter it using this command. If the host does not respond quickly enough, there might be a need to lengthen this time.
(Optional) Configure the TTL threshold:
```
 (interface)  ip multicast ttl-threshold   ttl  
```
This sets the maximum number of hops that a packet will be forwarded. If a multicast packet is received with a TTL greater than the TTL threshold, it is forwarded by the interface; otherwise , it is dropped.
(Optional) Disable fast switching for IP multicast packets:
```
 (interface)  no ip mroute-cache  
```
It is useful to do this when you want to log debug messages for multicast packets.
(Optional) Enable CGMP support:
```
 (interface)  ip cgmp  
```
This command enables CGMP support to communicate multicast forwarding information with Cisco switches.
(Optional) Enable functional addressing for IP multicasting on Token Ring interfaces:
```
 (interface)  ip multicast use-functional  
```
This reduces the load on Token Ring clients that do not participate in the multicast, because they do not process packets that use the functional address.

Example

Figure 7-7 shows a network diagram for this example. In this example, multicast routing has been enabled, and the router has been set up to use the router with the address 199.9.9.2 as a rendezvous point. Each of the interfaces has been set up to use sparse-dense mode. Because an RP is configured, they will operate in sparse mode. Interface Ethernet 0 has also been configured to join multicast group 239.1.1.22, which allows the router to respond to pings to that group. IGMP parameters have been altered on interface Ethernet 1 to run Version 1 and query clients every 30 seconds.

Figure 7-7. Network Diagram for the IP Multicast Routing Example

graphics/07fig07.gif

  ip multicast-routing   ip pim rp-address 199.9.9.2   interface ethernet 0   ip address 1.2.2.1 255.255.255.0   ip pim sparse-dense-mode   ip igmp join group 239.1.1.22   interface ethernet 1   ip address 199.9.9.1 255.255.255.0   ip pim sparse-dense-mode   ip igmp version 1   ip igmp query-interval 30   interface serial 0   ip address 1.5.5.1 255.255.255.0   ip pim sparse-dense-mode   router igrp   network 1.0.0.0   network 199.9.9.0