0

Network:

10.1.1.2 and 10.1.1.3 in same network. 10.1.1.2 has VPN server running in bridged mode. 10.1.1.4 - ip address of VPN client. All machines have Debian 6 installed.

I'm trying to test multicast using iperf.

Listener:

iperf -s -u -B 224.1.1.1 -i 1

Sender:

iperf -c 224.1.1.1 -u -T 32 -t 3 -i 1

These combinations work fine:

  1. Listener on 10.1.1.2, sender on 10.1.1.3
  2. Listener on 10.1.1.2, sender on 10.1.1.4
  3. Listener on 10.1.1.3, sender on 10.1.1.3
  4. Listener on 10.1.1.3, sender on 10.1.1.4
  5. Listener on 10.1.1.4, sender on 10.1.1.3
  6. Listener on 10.1.1.4, sender on 10.1.1.4

These combinations fail:

  1. Listener on 10.1.1.3, sender on 10.1.1.2
  2. Listener on 10.1.1.4, sender on 10.1.1.2
  3. Listener on 10.1.1.2, sender on 10.1.1.2

So, I can't send any multicast packets from 10.1.1.2.

ifconfig of 10.1.1.2:

br0       Link encap:Ethernet  HWaddr 00:25:90:e3:1e:0c  
          inet addr:10.1.1.2  Bcast:10.1.1.255  Mask:255.255.255.0
          inet6 addr: fe80::225:90ff:fee3:1e0c/64 Scope:Link
          UP BROADCAST RUNNING PROMISC MULTICAST  MTU:1500  Metric:1
          RX packets:31968583 errors:0 dropped:0 overruns:0 frame:0
          TX packets:29453949 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0 
          RX bytes:3140243824 (2.9 GiB)  TX bytes:243929369766 (227.1 GiB)

eth0      Link encap:Ethernet  HWaddr 00:25:90:e3:1e:0c  
          inet6 addr: fe80::225:90ff:fee3:1e0c/64 Scope:Link
          UP BROADCAST RUNNING PROMISC MULTICAST  MTU:1500  Metric:1
          RX packets:2196690402 errors:0 dropped:0 overruns:0 frame:0
          TX packets:8254769825 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:344354922111 (320.7 GiB)  TX bytes:11202019887816 (10.1 TiB)
          Memory:dfb20000-dfb40000 

tap0      Link encap:Ethernet  HWaddr be:fb:ed:8a:ce:f9  
          inet6 addr: fe80::bcfb:edff:fe8a:cef9/64 Scope:Link
          UP BROADCAST RUNNING PROMISC MULTICAST  MTU:1500  Metric:1
          RX packets:163739 errors:0 dropped:0 overruns:0 frame:0
          TX packets:196107 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:100 
          RX bytes:14506871 (13.8 MiB)  TX bytes:28238942 (26.9 MiB)

Where the problem could be?

ov7a
  • 113
  • 6

2 Answers2

1

I'm not sure whether your multicast issue is at the datalink layer or network layer.

In the past, I did experience issues with Ethernet multicasting when the packets are IP; all IP packets with a multicast address were being discarded i.e. MAC addresses with the value of 1 set in the least-significant bit of the first octet.

However, I noticed multicast/broadcast frames for certain protocols were being pasted through such NLB, STP, ARP without issues. So I concluded that bridging Ethernet multicasting packets most likely looks at protocols to make bridging decisions related to Etherenet multicasting.

Since my network is structure in a way where Ethernet broadcast/multicast made no different, I simply broadcasted the multicast frames using ebtables.

Please read my previous post relating to this issue on the following link: http://forums.debian.net/viewtopic.php?f=5&t=111338

Alternatively, you can set bridge to not filter the frames and instead pass it to iptables to make the routing decisions.

supmethods
  • 275
  • 3
  • 14
  • Unfortunalely, this didn't help me. My installation of Debian 6 didn't even have ebtables installed, but after I installed and conifugred it, there were no changes at all. – ov7a Mar 20 '14 at 12:34
  • I was going to suggest try using Wireshark to see if traffic does get passed though. Great to hear that a reboot fixed the issue. – supmethods Mar 25 '14 at 01:20
0

Somehow after rebooting 10.1.1.2 multicast started to work as expected. Why did this happen is still mystery.

ov7a
  • 113
  • 6