Understanding the Multicast Address Dynamic Client Allocation Protocol MADCAP

DHCP is normally used to assign IP configuration information for unicast (or one-to-one) network communications. With multicast, there's a separate type of address space assigned from 224.0.0.0-239.255.255.255; addresses in this space are known as Class D addresses or simply multicast addresses. However, multicast clients also need to have an ordinary IP address: Clients can participate in a multicast just by knowing (and using) the multicast address for the content they want to receive.

How do clients know what address to use? Ordinary DHCP won't help because it's designed to assign IP addresses and option information to one client at a time. Realizing this, the Internet Engineering Task Force (IETF) defined a new protocol: Multicast Address Dynamic Client Allocation Protocol (MADCAP). MADCAP provides an analog to DHCP, but for multicast use. A MADCAP server issues leases for multicast addresses only. MADCAP clients can request a multicast lease when they want to participate in a multicast.

There are some important differences between DHCP and MADCAP. First, you have to realize that the two are totally separate. A single server can be a DHCP server, a MADCAP server, or both; there's no implied or actual relation between the two. Likewise, clients can use DHCP and/or MADCAP at the same timeā€”the only requirement is that every MADCAP client has to get a unicast IP address from somewhere.

Next, remember that DHCP can assign options as part of the lease process but MADCAP cannot. The only thing MADCAP does is dynamically assign multicast addresses.

^Jfcj The Windows Server 2003 online help has a comprehensive checklist that covers how to set up IP multicasting.

Was this article helpful?

0 0

Post a comment