Multicast filter does not avoid reply

From: gladston@br.ibm.com
Date: Wed Apr 13 2005 - 18:33:47 GMT-3


Although 3550 is configured with igmp filter, and show commands shows it is working, R3 reply to ping from R4.

R3-----(fa0/2)CAT2(fa0/3)-----R4

R3 and R4 are configured as PIM sparse-dense-mode. There is no RP. The goal is to test IGMP filtering.

nmcr4>ena
nmcr4#p
Protocol [ip]:
Target IP address: 239.2.2.2
Repeat count [1]: 99999
Datagram size [100]:
Timeout in seconds [2]:
Extended commands [n]:
Sweep range of sizes [n]:
Type escape sequence to abort.
Sending 99999, 100-byte ICMP Echos to 239.2.2.2, timeout is 2 seconds:

Reply to request 0 from 172.16.34.3, 4 ms
Reply to request 1 from 172.16.34.3, 4 ms

Group 239.2.2.2 does not appear on 3550:

CAT2#sh mac-address-table multicast vlan 34
Vlan Mac Address Type Ports
---- ----------- ---- -----
  34 0100.5e00.0128 IGMP Fa0/1, Fa0/2
  34 0100.5e01.0101 IGMP Fa0/1, Fa0/2

CAT configuration:

ip igmp profile 5
    range 239.2.2.2 239.2.2.2
!
vlan 34
!
interface FastEthernet0/2
 switchport access vlan 34
 switchport mode access
 no ip address
 spanning-tree portfast
 ip igmp filter 5

Isn't is strange? or may I missing something.

R3 is the IGMP query. I was wondering if it would cause this:
(as R3 is a router, 3550 forwards any multicast to the router, causing this; well, it would happens even if R3 was not the querier)

CAT2#sh ip igmp snooping querier v 34
Vlan IP Address IGMP Version Port
---------------------------------------------------
34 172.16.34.3 v2 Fa0/2



This archive was generated by hypermail 2.1.4 : Tue May 03 2005 - 07:54:57 GMT-3