© 2013 Cisco and/or its affiliates. All rights reserved. This document is Cisco Public Information. Page 30 of 70
Multicast Overview (Traditional and Converged Multicast)
Efficient and intelligent use of bandwidth is paramount, particularly with the advent of video, mobility, and cloud
technologies. It is also critical considering the surge in related one-to-many or many-to-many communication-
based applications. Multicast helps to fulfill the requirement of such bandwidth-intensive applications with its
inherent ability to replicate a single stream when and where necessary.
In today’s network, replication for wired clients is performed at the network switches. There are two methods that
multicast works in wireless. Either the controller replicates to the access points that have interested clients, or the
controller replicates one packet to the multicast group address to which all access points are joined, offloading the
replication to the multicast-enabled network infrastructure. Hence there is a duplication of one multicast source
stream: one for wired, one for wireless.
With the Cisco Catalyst 3850 and the wired/wireless convergence, there is only one multicast source stream in the
network. The Cisco Catalyst 3850 replicates this stream to both wired and wireless clients. This helps in reducing
the number of streams from the same source, thus conserving bandwidth and enhancing overall network
performance.
The wired multicast configuration is the same as that of existing Cisco Catalyst 3K Series switches. Refer to the IP
Multicast Configuration Guide for configuring IP multicast on the wired network.
Restrictions of IP Multicast Routing Configuration
The following are the restrictions for configuring IP multicast routing:
IP multicast routing is not supported on switches running the LAN Base feature set.
Multicast Flexlink is not supported on the switch.
Layer 3 IPv6 multicast routing is not supported on the switch.
Configuring Wireless IP Multicast on Cisco Catalyst 3850
There are two modes in which wireless multicast works on the Cisco Catalyst 3850. In the basic mode, the switch
replicates individual packets to only those ports where the access point has interested clients. In this mode, the
replication of packets occurs before the CAPWAP encapsulation is added to the packet, destined to each
interested access point. This increases the recirculation on the switch since each packet to each access point is
passed through the recirculation block. However, this leads to efficient use of bandwidth in the switch since the
replicated packets are only sent to those access points that have interested clients.
In multicast-multicast mode, the switch and all the access points connected to it join one unique multicast group
that is not used anywhere else in the network. The ingress source stream from the campus network is replicated
once and encapsulated in CAPWAP with the destination address of this multicast group (formed between the
switch and all access points). In this mode, the replication happens only once for the group - hence only one
recirculation for CAPWAP encapsulation - and this packet is transmitted to all the connected access points. In this
mode, number of packets being switched by the UADP ASIC is optimized over the cost of switch bandwidth. The
switch always uses the management interface IP address as the source for sending these CAPWAP-encapsulated
multicast packets. The access point decapsulates the outer CAPWAP header and sends the original multicast
packets as broadcast at the lowest data rate on all BSSID and radio.