IPv6 cheat-sheet, part 3: IPv6 multicast
Many IPv6 multicast addresses are well-known to software implementing IPv6, to simplify common routing needs.
Read moreMulticast is an IP addressing and delivery method that sends a single stream of packets to multiple receivers simultaneously by using group addresses, enabling efficient distribution of data to many hosts without requiring separate copies from the source. The technique uses reserved address ranges (224.0.0.0/4 for IPv4 and ff00::/8 for IPv6) and relies on membership protocols like IGMP and MLD to allow hosts to join groups, with routers replicating traffic only where group members exist. Poor multicast management causes difficult-to-diagnose network problems such as VLAN flooding, blocked membership messages, and failed routing adjacencies, but proper configuration of switch snooping, group address allocation, and traffic filtering ensures reliable delivery for critical functions like OSPF routing and IPv6 neighbor discovery.
Multicast separates the sender from the receivers by using a group address. The sender transmits once to that address, and the network delivers copies only to hosts that have asked to receive them.
Multicast delivery is typically sent over UDP and is not retransmitted by the network, so applications must tolerate loss or add their own recovery. Some addresses are reserved for infrastructure use. For example, OSPF uses 224.0.0.5 and 224.0.0.6 on IPv4, and IPv6 relies on multicast for neighbor discovery.
Without IGMP snooping or MLD snooping, a Layer 2 switch treats multicast like broadcast and floods it to every port. Symptoms include saturated access links and slow hosts that never joined the group.
When snooping is on but no querier exists in the VLAN, membership entries age out. Streams then work for a few minutes and stop, a pattern that repeats until the hosts send new reports.
Protocol Independent Multicast checks that traffic arrives on the interface used to reach the source. Asymmetric routing or a missing route to the source makes routers drop packets, so receivers see no stream.
Multicast packets sent with a TTL of 1 never cross a router. An application that works inside one subnet then fails between sites until the TTL or hop limit is raised.
Firewall rules often permit unicast but drop group addresses or IGMP and MLD messages. Routing adjacencies that rely on multicast hellos, such as OSPF, then fail to form.