device immediately stops forwarding traffic to the interface. This feature requires the entire VLAN be
configured for IGMP V3 with no IGMP V2 clients. If a client does not send a report during the specified
group membership time (the default is 260 seconds), that client is removed from the tracking list.
Every group on a physical port keeps its own tracking record. However, it can only track group
membership; it cannot track by (source, group). For example, Client A and Client B belong to group1
but each receives traffic streams from different sources. Client A receives a stream from (source_1,
group1) and Client B receives a stream from (source_2, group1). The device still waits for the
configured leave-wait-time before it stops the traffic because these two clients are in the same group. If
the clients are in different groups, then the waiting period is not applied and traffic is stopped
immediately.
To enable the tracking and fast leave feature for VLAN 20, enter the following commands.
device(config)#vlan 20
device(config-vlan-20)#multicast tracking
Syntax: [no] multicast tracking
The membership tracking and fast leave features are supported for IGMP V3 only. If any port or any
client is not configured for IGMP V3, then the multicast tracking command is ignored.
Enabling fast leave for IGMP V2
When a device receives an IGMP V2 leave message, it sends out multiple group-specific queries. If no
other client replies within the waiting period, the device stops forwarding traffic. When fast-leave-v2 is
configured, and when the device receives a leave message, it immediately stops forwarding to that port.
The device does not send group specific-queries. When fast-leave-v2 is configured on a VLAN, you
must not have multiple clients on any port that is part of the VLAN. In a scenario where two devices
connect, the querier device should not be configured for fast-leave-v2 because the port might have
multiple clients through the non-querier. The number of queries, and the waiting period (in seconds) can
be configured using the ip multicast leave-wait-time command. The default is 2 seconds.
To configure fast leave for IGMP V2, enter the following commands.
device(config)#vlan 20
device(config-vlan-20)#multicast fast-leave-v2
Syntax: [no] multicast fast-leave-v2
Enabling fast convergence
In addition to sending periodic general queries, an active device sends general queries when it detects
a new port. However, because the device does not recognize the other device's port up event, multicast
traffic might still require up to the query-interval time to resume after a topology change. Fast
convergence allows the device to listen to topology change events in Layer 2 protocols such as
spanning tree, and then send general queries to shorten the convergence time.
If the Layer 2 protocol cannot detect a topology change, fast convergence may not work in some cases.
For example, if the direct connection between two devices switches from one interface to another, the
rapid spanning tree protocol (802.1w) considers this optimization, rather than a topology change. In this
example, other devices will not receive topology change notifications, and will be unable to send queries
to speed up the convergence. Fast convergence works well with the regular spanning tree protocol in
this case.
To enable fast-convergency, enter the following commands.
device(config)#vlan 70
device(config-vlan-70)#multicast fast-convergence

Enabling fast leave for IGMP V2

FastIron Ethernet Switch IP Multicast Configuration Guide 25
53-1003085-02