Hope all is well…have a quick question, I have 3 Xconn into an exchange what I am trying to accomplish is I want to evenly distribute the traffic across all 3 Xconns and I was thinking I would be able to utilize the
ip mroute [ip-addr mask | ip-prefix] [next-hop | nh-prefix] [route-preference]
command to influence which Xconn to take. I was trying to find more info on this and the command reference on Cisco site wasn’t helpful. Does anyone know where I can get more informaiton on this command?
The ip mroute command is a command specifically used to create a static route in the multicast routing table. More information about multicast routing can be found here:
I don’t believe that this command will help you in what you want to achieve. From my understanding, you want to load balance your traffic across three links to the exchange. Solutions to this issue depend upon the nature of the links. Can you tell us some more about them? Are they layer 3 links? Are you exchanging any routing information with the exchange via these links through a routing protocol? Also, because different people use different terms in different ways, can you explain what you mean when you say Xconn? Is that a cross-connect?
Thank you for your response…hmm ok let me see if I can do this without confusing everyone.
So currently we have 3 cross-connect (Xconn)
my side 192.168.164.116…exchange side 192.168.164.117
my side 192.168.168.192…exchange side 192.168.168.193
my side 192.168.224.34…exchange side 192.168.224.34
On the exchange side all 3 Xconn terminate on one switch, same on my side
The exchange statically flood the groups to their edge route and all unicast routes are learned over the 192.168.224.34. So what I am doing is I take the exchange side of the Xconn and use that as the RP…so for example
ip pim rp-address 192.168.164.117 prefix-list groups-ULTRA-A1
ip pim rp-address 192.168.168.193 prefix-list groups-ULTRA-A2
ip pim rp-address 192.168.224.34 prefix-list groups-ULTRA-A3
so now the return path is evenly distributed over all 3 Xconn. WE work together over which group we want to be learned over which Xconn and update our ACL on ours accordingly
So now what is happening is the Exchange is going to stop statically flooding and start to use RPs. We peer with them via BGP, some of the RPs I can see being advertised on each Xconn
I think I have an idea how to do this, I was thinking I would have to put in static routes to force the return path over the Xconn I want them to. Does this sound feasible?
Thanks for the clarification. When you have multiple equal-cost paths, by default, all multicast routing techniques will choose the PIM neighbor with the highest IP address. You can, however, configure multicast so that it dynamically advantage of multiple equal-cost paths using what is known as IP Multicast Load Splitting across Equal-Cost paths.
There are two methods to employ this:
Enable ECMP multicast load splitting based on source address, source and group address, or based on source, group, and next-hop address. After the equal-cost paths are recognized, ECMP multicast load splitting operates on a per (S, G) basis, rather than a per-packet basis as in unicast traffic.
Consolidate two or more equal-cost paths into a GRE tunnel and allow unicast routing to perform the load balancing.
I believe the first option is cleaner and more appropriate for your scenario. Take a look at the following Cisco documentation for more details of its operation as well as implementation.