6. Auto-RP Redundancy
Now in this video we’ll see the configuration of auto RP, multiple rps. Now the main advantage we get with the multiple rps is it’s going to provide us the redundancy. And also, if you want, we can also do something called load sharing. So we are going to practically verify how to configure multiple rps by using the cisco proprietary method. Now, if you just get back to some of the basic things, what we have learned. Now if you want to advertise the RP information, probably we need to configure two things like candidate rps. Now candidate RP are the possible routers which are advertising themselves as an RP. And then we have something called mapping agent. Mapping agent is something providing the information to the rest of the domain about the RP information.
Now all the routers must know who is the RP and our mapping agent is the one who is providing the information. Now there might be some scenarios where you may have an RP configured like in this scenario, let’s say I got an RP here, this router, and this RP goes down. Now, if there is no RP, if the RP goes down, in that case, probably the clients, let’s say I have a source sending a multicast traffic, will not be able to send Pim register messages. And also the receivers, I have a receiver here, will not be able to send Pim join messages. So RP is something mandatory if you’re using Pim sparse mode. And if I go with a single RP, then there might be some scenarios where if that RP goes down, it’s going to be a single point of failure.
And that is something we really don’t want. So we always want to have redundancy for the RP. Now in those kind of scenarios, we can configure multiple rps. Like in this scenario, I can configure the router one and router three as an rps. Now if you configure multiple rps, probably they will address themselves as an RP. And then the marting agent is going to learn the information about the RP from multiple routers. It’s going to select anyone as an RP based on the highest IP address. Let’s say in my scenario, the IP address of the router three is higher. So it’s going to decide that router three will be an RP and it’s going to propagate to the rest of the domain that router three is an RP.
Now, in case if this RP goes down due to some reason, the mapping agent is going to provide or propagate the information about router three is an RP. Now, at a time we are just using only one RP. If you configure multiple RP, only one RP is going to work. So we cannot use the same RP, two different routers as an RP for the same group. But we can do some load sharing. In some scenarios. When it comes to load sharing, what I can do is let’s say I got two different groups and these are the two different groups. We are using the multicast traffic and we can configure for load sharing, I can configure the router two is an RP. For this particular group, I can write in an acl and I can write another acl which is going to match this group address, 224 three.
And we are going to make the router Three as an RP. Now, if you configure separate rps for separate groups, it’s going to provide us the load sharing option where the RP is going to provide information that if anyone want to receive the multicast traffic for 224 two, it’s going to register with a Router Two. And the receivers or the sources who are sending the multicast traffic or receiving the multicast traffic on 224 three, they will register themselves with RP. Now, in this way we can reduce the extra overhead on a single router because if you make any one router as an RP, probably it’s going to add some more extra overhead. Now we can do something called load sharing.
Now probably in this video we’ll get into some more in detail with some configurations in the command line. So let’s go ahead and verify the same in our command line. Now, for verification, I’m going to use the same routers which we use in all our multicast labs. For the first thing, there is an igp which is pre configured here and I’m using osp of area zero and all the routers are forming the neighbor ship and the routes are coming. And then I’m going to enable IP Pim sparse mode here. So I’m going to start the configuration from the second step. And here I did the osp of configuration already because I don’t want to do that again here. And if you go and verify on the router four, if I show IP osp of neighbor.
Now I can see on the router four there are three neighbors. And if you verify the routing table, I’m able to see the route from him, from all the routers, from the router Five, router One, router Three and router Two. In a similar way, I’m going to check on the router Two as well, show IP osp of the neighbor. I got two neighbors and I’m going to enable IP pm’s passport on all the routers. So the first thing I’m going to say IP multicast routing. And then I’m going to enable on all the connected interfaces, interface S one by zero as well as interface S one by one. So I got the same interfaces on all the routers, so I’ll be using copy paste on all the routers. So almost on all the routers I got the same interfaces and on the router three as well.
And then on the router four, on the router four also I’m using the same interfaces. And on the router four I got one more extra interface that is S one by two. And on the router five. On the router five, I just got only two interfaces. One is s one by zero and s one by one. So I’m not enabling on s one by one here. So even if you enable, it doesn’t make difference because anyway, I’m not using s one by one here. Now, once we configure this pass mode, now in this scenario, what we’ll do is we will try to make multiple rps. So the first step we’ll see with multiple rp’s here. So I got my osp of configuration verification. If you want to verify the neighbors, that’s a good practice.
Always once you are done with Pim, verify the neighbors. On the router four, I should see three neighbors and on the router two, I should see how many neighbors. Two neighbors according to my topology show IP pim neighbors. Now the next thing is I want to configure the multiple rps here. Now we are going to configure two different routers and RP. So the router one will be router two and router three will be an RP and the router four will be a mapping agent. And we are going to use loopback zero interfaces for the RP. And we need to ensure that all the devices should be able to listen to that auto RP announcements. And to make that possible, we need to configure auto RP listener. So we have seen this in the previous classes.
Like by default, the auto RP uses 224-0139 and 40 multicasters. Each and every router has to listen on this multicaster addresses in order to learn about the RP information. Now, if you configure this command Ipm auto RP listener, it will allow the routers to listen to the advertisements. So let’s go to the configuration here. And my requirement here is I’m going to configure the router two is an RP, as well as router three, also will be an RP. So I’m going to configure two different rps so that anyone will be the RP here. And by default it’s going to select the RP based on the highest type of. And we are going to configure the router four as a mapping agent. So let’s configure this. So first I’ll start with the mapping agent configuration.
So on the router four, we are going to say router four will be an RP. So first thing I’m going to use the loop back interface. So I’m advertising the loop back interface in my ospf domain on the router four is having one loop back to 40 network. It’s pre configured in my topologies. So if you want to verify, we can verify which show IP interface pre. I got the loop back zero interface. It’s not present in the diagram, but it’s pre configured in my router here. And the next thing I’m going to add is I’m the RP. So I’m going to send the discovery messages and the loop bag zero and the scope. Let’s say I’m using ten the tdl value. So PM is ignored here because on that particular loop bag you must enable this pass mode.
That’s mandatory. And once I enable this, it’s not going to send any warning message. So I’m going to configure the Router Four as a mapping agent, which is going to send the discovery messages. And that particular interface is advertised in igp as well. Now the next thing we are going to configure the Router Two and Router Three will be an RP. So the first thing, I’ll start with Router three. And I’ll say that IP pim. And before I do this again, I need to ensure that I’m advertising the loopback interface on the Router three. I got a loop back interface here and then that particular loopback interface also should be enabled with pass mode. And then I’m going to say Send RP announcement, send RP announce loopac zero and the scope is ten.
So I’m going to use the same commands on the Router two, router Two. Also I’m going to configure the RP. Let me analyze the interfaces. So it’s a slash 24 and an interface loadbag zero sparse mode. And then I need to say IP pim or send RP announce. I’m going to make loopag zero as an RP. Now, once we configure the Router Two and Router Three as an RP, now in in this scenario I configure the Router Two, router Three as an RP for I didn’t mention any specific group, which means these are the rps for all the multicast groups. Now to verify this, we can go to Router router Four, which is my mapping agent. Let me write down here so that when we are moving so there’s an RP here, which is 120 loop back and the Router 13 one, that is also an RP.
And the Router four is a mapping agent. So that’s 41. Now in my scenario, if I use Show IP pimrp, IP mapping, Show Ipvmrp and then mapping, so there is a space between them. So if you go with the Show Ipvmrp mapping on the Router four. Now you can see this. Router Four says it’s a mapping agent and it’s going to learn from 2240 zero and RP is 31. Now, If I Try To Go To Router One On The Router One, if I Go show ipim rpm Mappings. The router One also learns About The RP. That Is Router Three. And now here actually there are two different RP adasmins coming from two different routers. And this advertisement will be mapping agent will see that there are two rps and it’s going to decide based on the highest IP address.
Now in my scenario, 13 one is the highest iprs. So the Router three is going to be decided as an RP. Now here, the Router Two cannot become an RP here unless the Router Three is down. Now for further verification, I can simply go to interface loopback zero. I’m going to shut down this router three interface, assuming that my router three goes down. Now I’ll go to Router four. I’ll use some command called clear IP, Pim RP Mappings. So just to refresh the mapping information, and if I use show IP Pim RP Mappings. And now I should see that router phone should learn the RP has to be router two. If you want to enable these messages, you can always go with debug ipm auto RP commands. To listen to these messages. You can always use that.
So I’m going to clear the mapping. Let’s see some of the debug commands as well. I can see the RP auto RP build Discovery packet. I’m expecting some messages. It takes some time. So let me use this clear IP mappings on all the routers. So maybe I’ll also do the same thing on all the routers. So just to clear it so that it can come faster. So I’m going to clear IP RP mappings on all the routers. Let’s do it on the Router One as well. On the Router Three. Because when you’re verifying some information, either you have to wait for the mappings to come up, or you can simply use clear Ipm rpm mappings. Now, here, if you try to see here, I’m not learning the RP information because actually, I forward one command here to configure.
That is PIMAT or placement. Now I just forwarded this command to configure this. I thought I configured that because if you just get back to the previous videos we have seen that by default by default. To learn about the arp information they must register to this addresses. And that can be done either by running a sparse mode or by using an auto. org prisoner. Now, in my scenario, I’m going to use Auto or listener. Here. Let me just quickly configure this on all the routers. ippm auto or blizzard. And I’ll do this on all the routers. So I’m going to configure on the Router One and then on the Router Two. On the Router Three and then on the Router four. Finally on the router file. Now. I did the shutdown command.
If you just observe here. Once I enable the debug command debug ipm auto RP command, then I cleared the RP mappings. I can see. I’m receiving RP announced from twelve one this time. And the reason is it’s building the mapping entry for this, that for all the multicast group purposes, the RP is twelve one. Okay? And then it is sending a discovery messages to all the router, all the interfaces saying advertising the RP information. Now, if we just get back to the Router One. Now, the Router One should say this router Two is selected as an RP. And the reason is the Router Three interfaces we did shut down. So let me make this interface back to up the low back interface.
And once this interface comes back to up now the Router Four should listen some RP messages from the Router Three as well once the interface comes back to up because the Router Three also will advertise the information. So you can see it’s receiving some RP announced from 120 it should get from router three as well. So I just did the clear Ipm RP mapping command just to remove all the RP information on the mapping agent. And then I’m expecting some RP announcement messages. Now I can see here I’m receiving RP announcement from the Router three. I can see the Router three is routing itself as an RP here. And if I go and verify show ipp RP mappings, I can see the RP information is 3001.
And if I go and check on the remaining routers on the router one, the router one will see the Router three as an RP. Now, in case if you configure multiple rps, probably the candidate RP will be decided based on the highest ipos. Now this way if any one of the RP fails, still the mapping agent is going to decide another router as an RP and it’s going to propagate that information to the rest of the domain. Now in this way we are getting redundancy option where if RP fails, we got a backup RP, but we cannot use both the rps at the same time for the same multicast group purchases. Now, if you want to do some load sharing, we will get into that in a continuation of this video. Probably in the next video we’ll continue with the load sharing as well.
