Quanta Cloud Technology QuantaMesh QNOS5, Руководство по настройке

Страница: 223 / 256

223
strips off the IP encapsulation and forwards the packet as an IP Multicast packet. This process of
encapsulating multicast packets in IP is called tunneling.
8.1.8. IGMP
The Internet Group Management Protocol (IGMP) is used by IPv4 systems (hosts, L3 switches, and routers) to
report their IP multicast group memberships to any neighboring multicast routers. The switch performs the
multicast router role of the IGMP protocol, which means it collects the membership information needed by
the active multicast routing protocol.
The switch supports IGMP Version 3. Version 3 adds support for source filtering, which is the ability for a
system to report interest in receiving packets only from specific source addresses, as required to support
Source- Specific Multicast [SSM], or from all but specific source addresses, sent to a particular multicast
address. Version 3 is designed to be interoperable with Versions 1 and 2.
8.1.8.1. IGMP Proxy
IGMP proxy enables a multicast router to learn multicast group membership information and forward
multicast packets based upon the group membership information. The IGMP Proxy is capable of functioning
only in certain topologies that do not require Multicast Routing Protocols (i.e., PIM-DM and PIM-SM) and have
a tree-like topology, as there is no support for features like reverse path forwarding (RPF) to correct packet
route loops.
The proxy contains many downstream interfaces and a unique upstream interface explicitly configured. It
performs the host side of the IGMP protocol on its upstream interface and the router side of the IGMP
protocol on its downstream interfaces.
The IGMP proxy offers a mechanism for multicast forwarding based only on IGMP membership information.
The router must decide about forwarding packets on each of its interfaces based on the IGMP membership
information. The proxy creates the forwarding entries based on the membership information and adds it to
the multicast forwarding cache (MFC) in order not to make the forwarding decision for subsequent multicast
packets with same combination of source and group.
8.1.9. MLD Protocol
MLD, Multicast Listener Discovery is a protocol for IPv6 multicast router. It is to discover the presence of
multicast listener on the local network and it is also used to discover which multicast packets are of interest
to neighboring nodes. MLD is a sub-protocol of ICMP v6, Internet Control Message Protocol version 6 ,and
MLD messages are a subset of ICMP v6 messages.
Router use MLD to learn whether multicast group members of a group are present in their directly
connected network. Each host in the multicast network sends MLD join report to join the multicast group
and send MLD Leave message at any time as IGMP protocol act. The Multicast packets are delivered to a
group using IPv6 unicast packet.
MLD v1 (RFC 2710) is equivalent to IGMP v2. MLDv2 (RFC 3810) is equivalent to IGMP v3 (MLDv2 is fully
backward compatible with MLD version 1). All MLD messages are link-local with a hop limit of 1, and they
all have “router alter option” set.