Configuring PIM Sparse Mode

1. From "Flood and Prune" to "Explicit Join"

Before we dive into PIM sparse mode (PIM-SM), it's helpful to briefly mention its counterpart, PIM dense mode (PIM-DM). PIM-DM uses a "flood and prune" or "push" model. It floods multicast traffic to every PIM-enabled router in the network and then relies on routers that don't need the traffic (because they have no interested receivers) to send "prune" messages back to stop the flow. While simple, this is highly inefficient, especially in networks where multicast receivers are few and far between—a "sparsely" populated environment.

PIM-SM was designed to solve this. It operates on an "explicit join" or "pull" model. No multicast traffic is forwarded unless a router has explicitly requested it on behalf of a receiver. This is far more scalable and is the standard for multicast in most enterprise networks.

The central component that makes this explicit join model work is the Rendezvous Point (RP).

2. The Role of the Rendezvous Point (RP)

Think of the RP as a designated meeting point in the network for a particular multicast group.

• Sources "register" with the RP: When a source starts sending traffic, its first-hop router (FHR) informs the RP.
• Receivers "meet" at the RP: When a receiver wants to join a group, its last-hop router (LHR) sends a join message towards the RP.

This coordination by the RP allows PIM-SM to build two types of distribution trees:

1. Shared Tree (*,G): This is the initial path built from the RP down to the receivers. It's called a shared tree because multiple sources for the same group can use it. The notation (*,G) signifies a tree for any source (*) to a specific group (G).
2. Source Tree (S,G): This is a more optimal path directly from a specific source (S) to a specific group (G). After a receiver starts getting traffic via the shared tree, its LHR can choose to build a more direct, lower-latency path by creating a source tree. This process is called SPT (Shortest Path Tree) switchover.

The video below provides an excellent conceptual overview of these components, contrasting PIM-DM with PIM-SM and illustrating the function of the RP and the two tree types.

This diagram illustrates the operational flow you just saw in the video, from neighbor discovery to building the final shortest path tree.

3. PIM Sparse Mode Configuration

Now, let's get practical. Configuring PIM-SM involves three fundamental steps on all routers participating in the multicast domain:

1. Enable Multicast Routing Globally: This is the master switch for multicast on the router.
2. Enable PIM on Interfaces: You need to tell the router which interfaces will participate in PIM.
3. Define the Rendezvous Point: All routers must agree on the location of the RP.

Let's break down the commands.

Step 1 & 2: Base Configuration

These commands are the same for all PIM routers in your domain.

! On every multicast router
!
! 1. Enable multicast routing globally
Router(config)# ip multicast-routing

! 2. Enable PIM sparse mode on participating interfaces
Router(config)# interface <type/number>
Router(config-if)# ip pim sparse-mode

You must enable PIM on all Layer 3 interfaces that will carry multicast traffic, including those facing sources, receivers, and other routers.

Step 3: Configuring the Rendezvous Point

This is the most critical part of the configuration. How do all the routers learn the address of the RP? There are two primary methods you need to know for the ENCOR exam: Static and Auto-RP (dynamic).

Method 1: Static RP Configuration

This is the simplest method. You manually configure the IP address of the RP on every single router in the PIM domain.

• Best Practice: The RP address should be an IP on a loopback interface of the designated RP router. Loopbacks are always up, making them more resilient than physical interface IPs.

The configuration is straightforward.

! On ALL routers in the PIM domain, including the RP itself
!
Router(config)# ip pim rp-address <RP_LOOPBACK_ADDRESS>

You can optionally use an access list to specify that this RP is only for a certain range of multicast groups. If no ACL is specified, the RP serves all groups (224.0.0.0/4).

The IPCisco article provides a clear, step-by-step example of a full static RP configuration, including the underlying OSPF routing.

To see this in action, the following video segment demonstrates the static RP configuration and verification.

Method 2: Dynamic RP with Auto-RP (Cisco Proprietary)

Static configuration is simple but doesn't scale well and has no redundancy. If the RP fails, multicast stops working. Auto-RP is a Cisco-proprietary dynamic mechanism to solve this. It introduces two roles:

• Candidate-RP (C-RP): A router that wishes to be an RP. It announces its availability.
• Mapping Agent (MA): A router that listens for C-RP announcements, elects an RP, and advertises the group-to-RP mapping to all other PIM routers.

The process has a "chicken-and-egg" problem: routers need to learn the RP via multicast messages, but they need an RP to do sparse-mode multicast. This is solved by using ip pim sparse-dense-mode on the interfaces, which allows the Auto-RP control plane messages (on 224.0.1.39 and 224.0.1.40) to be flooded using dense mode, while data traffic uses sparse mode once the RP is known.

Here is a summary of the Auto-RP configuration from the official Cisco guide.

Now, let's watch a demonstration of this configuration.

4. Verification

Once configured, you must verify that PIM is operating correctly. Here are the most important commands:

For a very deep dive into the PIM state machine, including the register process and tree building, the "Decoding Packets" video series is an excellent resource. The following segments show what happens when only a receiver is present, and when only a source is present. This is great for understanding the underlying mechanics.

Conclusion

In this lesson, we have moved from local group management with IGMP to enterprise-wide multicast distribution with PIM sparse mode. You have learned how PIM-SM uses an explicit-join model, centered on the Rendezvous Point, to build efficient multicast trees.

Key Takeaways:

• PIM Sparse Mode is a "pull" model, sending traffic only where it is requested, making it highly scalable.
• The Rendezvous Point (RP) is a central coordinator for sources and receivers.
• PIM-SM uses two trees: an initial Shared Tree (*,G) from the RP, and a more optimal Source Tree (S,G) built directly from the source.
• Static RP configuration (ip pim rp-address) is simple but requires manual configuration on every router.
• Auto-RP is a dynamic, Cisco-proprietary method using Candidate-RPs and Mapping Agents to automate RP discovery.
• Verification is key: use show ip pim neighbor, show ip pim rp mapping, and show ip mroute to troubleshoot and confirm operation.

Preview of the Next Lesson:

Now that we can deliver specialized traffic like video and voice streams efficiently using multicast, the next logical step is to ensure that this delay-sensitive traffic receives preferential treatment as it traverses the network. In our next lesson, we will analyze QoS mechanisms for classification, marking, queuing, policing, and shaping, which are essential for guaranteeing the performance of real-time applications.