SRX

Hi,

 

I have attached an overview of my network to help explain.

 

At 172.31.255.102 I have two video encoders which output the video in multicast stream. There are decoders at 172.31.255.108,109,151,152 and 153.

 

I now need to install a video encoder at 172.31.255.107 and have a decoder at 172.31.255.108.

 

My current config is:

172.31.255.102:

pim {
    rp {
        local {
            address 172.31.255.102;
        }
    }
    interface all {
        mode sparse;
    }
    interface ge-1/0/0.0 {
        mode sparse;
        version 2;
    }
}

 

All other routers:

pim {
    rp {
        static {
            address 172.31.255.102;
        }
    }
    interface all {
        mode sparse;
    }
}
igmp-snooping {
    vlan all;
}
igmp

 

I have now added the following to 172.31.255.107

pim {
    rp {
        static {
            address 172.31.255.102;
        }

       local{

       address 172.31.255.107
    }
    interface all {
        mode sparse;
    }
}
igmp-snooping {
    vlan all;
}
igmp

 

and then added pim { rp { static { address 172.31.255.107 } } } to all other routers

 

However I'm not able to get any video from the 172.31.255.108 router.

 

What tests can I do to try and diagnose this? Or am I missing some config?

 

Thanks

 

Andrew.

Hello,

I'm a little confused by your RP design. Do you intend to have two active RP's for the same multicast groups at the same time?

You don't need a local-RP on every machine that has a multicast-source.

As for troubleshooting, you need to verify these:

1. PIM neighborships are formed on every layer-3 interface in the path.
2. Status of *,G and S,G entries
3. I see IGMP-snooping too, make sure that the mrouter ports are learned fine (wherever applicable).
4. Make sure there's no RPF check-failures.

HTH,
ankit

Ankit,

 

Thanks for your response.

 

I'll be honest, I got this working with the RPs more by testing than understanding last time (not great, I know, but I was short on time).

 

Can you explain a bit please? Are you saying I should just have the one RP point?

 

Andrew

>>Are you saying I should just have the one RP point?

This could start a philosophical debate. 🙂
Now without being politically correct, having just one RP in the whole network would well. In this case, you should be able to ale it work with just .102 acting as the RP in the network. You need to make sure that these things are in place:

1. RP is reachable from every other router (looks like this is already true otherwise you would've had issues earlier as well).
2. There should be no reverse-path-failures.
(and the other points that I'd mentioned above).

This link has a great "checklist" for the things you should consider to get multicast running:
http://www.juniper.net/techpubs/en_US/junos13.2/topics/task/configuration/routing-protocol-multicast-security-configuring.html

Note that you're using static-RP and that IGMP automatically gets enabled when you enable PIM.

Please let me know how it goes.

ankit

Ok, I'll go through that document, thank you.

 

I don't know if it's worth mentioning, but I did try just using the new address (239.0.0.100) without any changes and I didn't have any success with it.

 

Currently I have streams running on:

239.0.0.1

239.0.0.2

239.0.0.3

239.0.0.4

239.0.0.5

239.0.0.6

239.0.0.7

239.0.0.8

 

And now I've added 239.0.0.100 for this new stream

This is good information. I'll PM you my email address and we can probably iron this out offline.

ankit

In case anyone finds this thread with the same issue, this was resolved by removing the duplicated static RP which had been configured.