Multicast problem on cisco831(help!!!)

[bigeric]

Dear Sir,

I have one Cisco 837 for testing my multicast application. The setup is as follows

Host A --- eth1 -- Cisco831 -- eth0 -- Host B

Both host A and B listen to the same multicast group (239.0.1.1) for mutual communication. This means host A send message to host B via 239.0.1.1 and host B also send message to host A via 239.0.1.1

The configuration is basically simple

ip multicast-routing

interface ethernet 0

ip pim spare-dense-mode

interface ethernet 1

ip pim spare-dense-mode

However, my application doesn't work... is there any thing that I can try to shoot? is there any multicast testing program so that I can try to isolate the problem in application layer

thx

Eric

[Harold Ritter]

Can you post the output of the "show ip mroute" command.

Thanks,

[Harold Ritter]

BTW: you should be able to do a ping 239.0.1.1 to generate multicast traffic from one workstation to another.

Hope this helps,

[bigeric]

This is the setup

Host A -- eth0 -- Cisco831 -- eth1 -- Host B

This is the configuration

##############

ip cef

ip multicast-routing

no ip ips deny-action ips-interface

!

no ftp-server write-enable

!

interface Ethernet0

ip address 192.168.1.1 255.255.255.0

ip pim sparse-dense-mode

!

interface Ethernet1

ip address 202.67.194.251 255.255.255.224

ip pim sparse-dense-mode

ip igmp join-group 225.2.2.2

duplex auto

!

interface Ethernet2

no ip address

shutdown

!

interface FastEthernet1

no ip address

duplex auto

speed auto

!

interface FastEthernet2

no ip address

duplex auto

speed auto

!

interface FastEthernet3

no ip address

duplex auto

speed auto

!

interface FastEthernet4

no ip address

duplex auto

speed auto

!

ip classless

!

pim bidir-enable

ip pim autorp listener

#####################

I have attached the show information

From host A, I can ping the join group address 255.2.2.2 of eth1 . However in host A, the multicast application doesn't work with host B. (The application replies on both host A and host B sending message to each other via multicast group address 236.3.0.1, 236.3.2.2 etc)

[Harold Ritter]

The mroute entries look good. It seems that host A and B are receivers and sources for both multicast group (236.3.0.1 and 236.3.2.2).

I would check on the application side.

Hope this helps,

[bigeric]

Dear hritter,

My application on hostA and host B send/listen to the same multicast group for communication, is there any issue ?

I have also tried another program call network text editor (based on multicast) but also failed....

Would you mind to share any multicast testing program so that I can check out if the router can really pass multicast ?

thx

[Harold Ritter]

It is actually pretty common to see multicast applications where multiple hosts act as both source and receiver for the same group.

Please get a look at Iperf as a multicast testing tool:

http://dast.nlanr.net/Projects/Iperf/

Hope this helps,

[Richard Burts]

Eric

Can host A and B do other things ok? Can host A and B ping each others unicast IP address? I am wondering if there might be some issue with configuration of address, mask, or default gateway that might be part of the problem?

HTH

Rick

[bigeric]

Dear Rick,

There is no problem on connectivity, host a and host b can ping each other. I just wonder if this is my application issue. However, i don't know how to isolate.. do you have any window-based multicast testing tools ?

thx

[Harold Ritter]

I just tought of something. Is it possible that the application is designed to run on the same subnet and that it uses a TTL of 1 by default? This would definitely be an issue.

Hope this helps,

[bigeric]

Dear Hritter,

Do you mean that the application may be with TTL of 1... so the multicast traffic can not pass through even one hop...

Thx for your multicast testing tools, but do you have any window-based one ?

Eric

[Harold Ritter]

That is correct. If the application generates ip packets with a TTL of 1, they would be received by the router, get decremented and then dropped. A TTL of 2 would be required in your case. TTL of 1 will keep packets on the directly connected network.

I think that it is probably the issue in your scenario because the mroute entries looked good.

The link I sent does have an executable for MS Windows. That is the one I use myself.

[Harold Ritter]

Have you had the chance to verify that your application was indeed using a TTL of 1 or to run a quick test using Iperf with a TTL of 2.

Thanks for the feedback,