IP multicast is a bandwidth-conserving technology based on the IP protocol in which a data intended for multiple receivers is efficiently transmitted with a single stream. IP multicast best serves streaming audio, video, and data applications such as software distribution or stock-quote broadcasts.
Take this example: A corporate officer needs to deliver a live video broadcast to the company’s employees. Doing so across an IP network using traditional IP protocols would require that the video feed be replicated once for every user. For a branch office across a wide-area network (WAN) link, 20 employees watching the video on their computers would result in the video stream being replicated 20 times across the WAN. The original data stream then must be replicated 20 times across the single WAN connection. Not only is this highly inefficient, but the WAN link would not be able to handle the sheer amount of traffic being transmitted.
IP multicast protocols allow users to request a particular multicast service (which can be audio, video, or data). The network then determines the source of the multicast transmission and routes the multicast stream to its destination. Multicast protocols ensure that only one copy of the stream is transmitted across any given link. Therefore, in the previous branch-office example, only one copy of the video broadcast would be transmitted over the WAN and then replicated to each user at the local-area network (LAN) level.
Multicasts relate to the concept of a group. The receivers who express an interest in receiving a particular data stream define the group. A receiver must join the group by using the Internet Group Management Protocol (IGMP), which requires that routers and switches must support IGMP to communicate with the receiver. Not until the receiver joins the group can she receive the data stream.
A multicast IP address identifies a group. The Internet Assigned Numbers Authority (IANA) assigns multicast IP addresses using the IP Class D address space 224.0.0.0–239.255.255.255.
The routers in a network are responsible for identifying the most efficient, shortest path from the multicast transmitter to its group of receivers. The Layer 2 LAN switches are responsible for receiving the singular transmission and replicating it to each of the registered receivers.
Of the multicast-related routing protocols, the primary protocol is Protocol Independent Multicast (PIM). The routers use the PIM protocol to make forwarding decisions. PIM determines the shortest path for each multicast group based on the IP routing protocol running on the router (such as Enhanced Interior Gateway Routing Protocol, EIGRP, or Open Shortest Path First, OSPF). Using Reverse Path Forwarding (RPF), routers use the existing unicast routing table to determine upstream and downstream neighbors and eliminate any loops.
PIM operates in three different modes: sparse, dense, and sparse-dense. The mode determines when the routers start forwarding packets to the group. Sparse mode employs a “pull” model, and dense uses a “push” model.
In sparse mode, packets are not transmitted through the network until a receiver specifically requests to join the group. In dense mode, packets are transmitted throughout the network without regard to registered users, and the network prunes transmission to parts of the networks where no receivers are registered. Sparse mode is more efficient in that traffic goes only where requested, whereas dense mode blasts everywhere initially. Sparse-dense mode is the third mode (developed by Cisco) in which the network intelligently determines which mode to use as opposed to configuring just one or the other.
Sparse mode is generally implemented for standard video and audio transmission in corporate networks. A core component of a sparse implementation is the rendezvous point (RP). The RP is the central clearinghouse for multicast transmitters and group receivers. Multicast transmitters register themselves with the RP, and the network forwards requests from group receivers to the RP. The RP then facilitates the transmission of the data stream to the receivers.
If the RP fails or disappears from the network, receivers cannot register with senders. Therefore, a network needs multiple RPs. Anycast-RP is a protocol based on the Multicast Source Discovery Protocol (MSDP) that facilitates the fault-tolerant implementation of multiple RPs.
Dense mode floods traffic to every corner of the network. This brute-force method only makes sense when there are receivers on every subnet in a network. After traffic begins flowing, routers then stop the flow of (or prune) traffic to subnets that have no receivers. Companies generally do not implement dense mode except for special circumstances because of its fire-hose approach to data-stream replication.
Although PIM makes sense in a corporate network, it requires that a network be planned and configured to handle the particular implementation of PIM properly. This setup is not practical across the Internet and shared IP networks for which there is no guarantee that multicast is configured consistently or properly. Multiprotocol Border Gateway Protocol (MBGP) allows the scalable advertising and transmission of MBGP routes through the Internet.
At-A-Glance—IP Multicast
Many applications in modern networks require that you send information (voice, video, or data) to multiple end stations. When you target only a few end stations, sending multiple copies of the same information through the network (unicast) causes no ill effects. However, as the number of targeted end stations increases, the harmful effects of duplicate packets becomes dramatic. Deploying applications such as streaming video, financial market data, and Internet Protocol (IP) telephony–based Music-on-Hold without enabling network devices for multicast support can cause severe degradation to a network’s performance.
</division><division> <title>What Are the Problems to Solve?</title>
Multicasting requires methods to efficiently deploy and scale distributed group applications across the network. Multicasting uses protocols that reduce the network load associated with sending the same data to multiple receivers and alleviate the high host/router processing requirements for serving individual connections.
</division><division> <title>Internet Group Membership Protocol</title>Internet Group Membership Protocol (IGMP) allows end stations to join what is known as a multicast group. Joining a multicast group is like subscribing to a session or service that uses multicast. IGMP relies on Class D IP addresses for the creation of multicast groups.
When a multicast session begins, the host sends an IGMP message throughout the network to discover which end stations have joined the group. The host then sends traffic to all members of that multicast group. Routers “listen” to IGMP traffic and periodically send queries to discover which groups are active or inactive on particular LANs. Routers communicate with each other using one or more protocols to build multicast routes for each group.
</division><division> <title>Multicast Distribution Trees</title>Multicast-capable routers create distribution trees that control the path that IP multicast traffic takes through the network to deliver traffic to all receivers. The two basic types of multicast distribution trees are source trees and shared trees.
With source trees (also known as shortest-path trees), each source sends its data to each receiver using the most efficient path. Source trees are optimized for latency but have higher memory requirements because routers must keep track of all sources.
Shared trees send the multicast data to a common point in the network (known as the rendezvous point [RP]) prior to sending it to each receiver. Shared trees require less memory in routers than source trees but might not always use the optimal path, which can result in packet delivery latency.
</division><division> <title>Layer 2 Multicast</title>
A Layer 2 switch forwards all multicast traffic, which reduces network efficiency. Two methods, Cisco Group Management Protocol (CGMP) and IGMP Snooping, mitigate this inefficient switch behavior.
</division><division> <title>Cisco Group Management Protocol</title>CGMP allows catalyst switches to make Layer 2 forwarding decisions based on IGMP information. When configured on switches and routers, CGMP ensures that IP multicast traffic is delivered only to ports that are attached to interested receivers or multicast routers. With CGMP running, any router receiving a multicast join message via a switch replies to the switch with a CGMP join message. This message allows Layer 2 forwarding decisions.
</division><division> <title>IGMP Snooping</title>IGMP Snooping improves efficiency by enabling a Layer 2 switch to look at Layer 3 information (IGMP join/leave messages) sent between hosts and routers. When an IGMP host report travels through a switch, the switch adds the port number of the host to the associated multicast table entry. When the switch hears the IGMP leave group message from a host, the switch removes the table entry of the host. IGMP requires a switch to examine all multicast packets and therefore should only operate on high-end switches.
</division><division> <title>Multicast Forwarding</title>In unicast routing, traffic moves from the source to the destination host. The router scans through its routing table for the destination address and then forwards a single copy of the unicast packet out the correct interface.
In multicast forwarding, the source sends traffic to several hosts, represented by a multicast group address. The multicast router must determine which direction is the upstream direction (toward the source) and which one is the downstream direction (toward the hosts). When there is more than one downstream path, the router chooses the best downstream path (toward the group address). This path might not be the same path chosen for a unicast packet. This process is called Reverse Path Forwarding (RPF). RPF creates distribution trees that are loop free.
</division><division> <title>Protocol Independent Multicast</title>Protocol Independent Multicast (PIM) is IP-independent and can leverage whichever unicast routing protocols populate the unicast routing table. PIM uses this unicast routing information to perform the multicast forwarding function. Although PIM is called a multicast routing protocol, it actually uses the unicast routing table to perform the RPF check function instead of building up a completely independent multicast routing table. It includes two different modes of behavior for dense and sparse traffic environments, dense mode and sparse mode.
PIM dense mode—. In dense mode, the multicast router floods traffic messages out all ports (referred to as a “push” model). If a router has no hosts or downstream neighbors that are members of the group, a prune message tells the router not to flood messages on a particular interface. Dense mode only uses source trees. Because of the flooding and pruning behavior, dense mode is not recommended.
PIM sparse mode—. PIM sparse mode uses an “explicit join” model. This model sends traffic only to hosts that explicitly ask to receive it. The router sends a join message to the RP.
Anycast RP provides load balancing, redundancy, and fault tolerance by assigning the same IP address to multiple RPs within a PIM sparse mode network multicast domain.
</division>