multicast, 2 NIC's

[Jst]

Hi,

here is the situation I have. In my computer I have 2 NIC's. Eth0 is for accessing the outer world (internet that is) and eth1 for the TV. The problem is I cannot get those two things to work at the same time. If I wish to watch TV I have to do "route add -net default eth1" and that makes the net inaccessible. I tried doing "route add -net 239.255.0.0/24 eth1" (the streams are on 239.255.0.x) but that does not help.

Thanks for all the help.

[MaSTeRMinD]

I'm bumping my head for two days with the same problem. I still don't have good enough solution. Trying with ubuntu 8.10.
When computer boots i cannot start multicast streams with vlans. If bring down all interfaces and keep the one with the multicast streams after a while they are fine. Then when bring up other interfaces all is OK. Even loopback is causing troubles.
I'm using this routing for the multicast
route add -net 224.0.0.0/4 dev eth0

You should use route add -net 224.0.0.0/4 dev eth1 and add
route add default gw router_IP

I tried the same setup on windows and everything works as it's supposed to work.

This is what I see in the log of VLC. I

Code: Select all

main debug: creating demux: access='udp' demux='' path='@239.20.20.101:10135' main debug: looking for access_demux module: 0 candidates main warning: no access_demux module matched "udp" main debug: TIMER module_Need() : 0.081 ms - Total 0.081 ms / 1 intvls (Avg 0.081 ms) main debug: creating access 'udp' path='@239.20.20.101:10135' main debug: looking for access module: 1 candidate access_udp debug: opening server=:0 local=239.20.20.101:10135 main debug: net: opening 239.20.20.101 datagram port 10135 main debug: Multicast group join request main debug: using access module "access_udp" main debug: TIMER module_Need() : 0.525 ms - Total 0.525 ms / 1 intvls (Avg 0.525 ms) main debug: Using AStream*Block main debug: pre buffering main debug: thread 2956172176 (input) created at priority 10 (input/input.c:370) qt4 debug: Updating the stream status: 3

and tcpdump at this time

Code: Select all

tcpdump: listening on eth0, link-type EN10MB (Ethernet), capture size 96 bytes 08:51:09.572603 IP (tos 0xc0, ttl 1, id 0, offset 0, flags [DF], proto IGMP (2), length 40, options (RA)) 20.20.20.99 > 224.0.0.22: igmp v3 report, 1 group record(s) [gaddr 239.20.20.101 to_ex { }] 08:51:18.664073 IP (tos 0xc0, ttl 1, id 0, offset 0, flags [DF], proto IGMP (2), length 40, options (RA)) 20.20.20.99 > 224.0.0.22: igmp v3 report, 1 group record(s) [gaddr 239.20.20.101 to_ex { }

I think that cause is this
net.ipv4.conf.default.rp_filter=0
net.ipv4.conf.all.rp_filter=0
Try to set them in /etc/sysctl.conf and then sysctl -p

[Jst]

Wow, fixed it. I checked the reverse path filtering as you suggested and it really was causing the problems.