1. PIM Sparse Mode – Operation
Now, in this video, we’ll see how pim’s password is going to operate in different steps. Now, before we go ahead, let’s quickly revise the pimp sparse mode. We have two different modes, a sparse mode and the Desk mode. Now, the sparse mode is much more scalable when compared with the desk form, because in case of dense mode, the one which we discussed in the previous sections, it’s going to work based on flood and prune behavior. So initially, the traffic is flooded out of all the interfaces. And then if you have a receiver who do not want to receive the multicast traffic, he’s going to send out approved messages. Now, the major drawback with the dense mode is it’s really not scalable. But in case of sparse mode, it is not going to do flooding.
Instead, it is going to send the multicast traffic only when it is requested form. So there’s a major difference between the sparse mode and the dense mode. So it uses explicit join. Explicit join is just like you need to tell that I want to receive the multicast traffic, then only the multicast traffic will be sent or else it will not send unless you ask for it. It uses both shit tree as well as source tree. We’ll see how it’s going to use Shadery and source tree more in detail. And probably the sparse mode is much more scalable solution when compared with a dense mode. Now, the complete dense mode sparse mode operation is going to happen in six different steps. Now, we’ll try to get into more in details on what are the different steps it’s going to do.
Now, the first step, the first step is it discovers the Pim neighbors. Now, the router, let’s say we have a router in router C, and these two are configured with a command called under the interface, we need to say I p PM pass mode, just like we did the configuration of dense mode. Now, once we enable this command on the interface, it is going to discover the neighbors by based on some multicasters to 240 00:13. Same like same like dense mode. And there is a periodical hello messages sent between these two neighbors for every 30 seconds and the default date time is 90 seconds. Now, the first thing they need to discover the neighbors. So like the the same thing happens between every two routers. They need to identify the Pim neighbors.
Now, once they discover the Pim neighbors, once we enable a multicasting and the Pim on all the routers. Now the next step is they need to decide the RP or every router should learn about the random point. Now, the RP is the root of the tree. We can say it is the root of the tree or a central point or the router which is going to provide the information about the source who is sending the multicast traffic and to the destination. So the information from source to destination is something provided by RP. We can say RP is a reference point for the complete root of the shades. If you can send now, now in order to send and this is the traffic, the source who is sending the multicast traffic should know about RP and the receiver who wants to receive the multicast traffic.
They should know who is RP so that they can register themselves with RP and RP will provide the information multicast traffic from source receiver. Now, this RP can be manually configured by using some commands. We can either specifically go and manually configure on each and every router that who is RP or we can advertise the RP information to all the routers by using some dynamic RP configurations. So we’ll be getting into that more in detail with the specific configurations in the later on sessions. Now, once the routers learns about the RP, now the next step is now the source. Now, the source who is sending the multicast traffic, they need to register themselves with RP. So that’s the next step. So RP listen to the source by using some Pim register messages. Now in this third step, now the source must register themselves with RP.
Now the source here is going to send something called Pim register messages and they should register with RP by sending some unicast Pim register RP messages. Now, if the RP receives or accept that particular message, it acknowledges back with a register stop message. Now, register stop message is just like a confirmation that okay, you are an RP, I’m going to agree, sorry, you are a source and the source is going to send that I am the source for this one. So multicast group address and this is my IP address, ten one or one. And the RP is going to maintain some S comma G entries in that saying that if any, if I have any receiver who wants to receive the multicast traffic for this group address, I’m going to send it back to ten one one and it’s going to write that in the routing table.
Now, at this point, only the source router and the RP knows about this S comma G entry. Now, the RP knows the entry for source and the group and the particular source knows this entry. Whereas all the remaining routers, they don’t know any information about the source and the group. But whereas in case of dense mode, every router simply gets the information of S commercial entries. But here it’s not like that. Now, the source has to register themselves with RP and they had to tell that I am the source for this particular multicast group and this is my IP address. Now again, there is something called rpf check happens here as well. Now the rpf check is done based on the register message it receives from the source.
And if the RP do not have a route back to the source on the receiving interface, it is not going to accept the registration messages or Pim register messages that’s something like the source is going to send back a Pim register message. Now, when they’re sending a Pim register message, it’s going to verify the reverse path forwarding check back to the source and it will ensure that on whatever the interface it is receiving, it’s going to match, it should match that RP of check. And if it matches, it’s going to accept the Pim register messages and it’s going to acknowledge back with the register’s top messages. If not, it’s not going to accept that. So again, the rpf check is also done here based on the same conditions again. Now, the fourth step, RP should also know about the destination as well.
Like here, just now we discussed that RP is the root of the tree. Now, the source has to register with some Pim register messages. Now, similar way if you have any receivers, like, let’s say I have a receiver here who wants to receive the multicast traffic for this particular group and this is the source address for this particular group, or the RP knows the information. Now, what the receiver will do is receiver is going to request or send an icmp join message to the nearest hop router. That is the first hop router. And what this particular router is going to do is it’s going to send back a join message to the RP saying that, hello RP, I have a receiver who wants to receive the multicast traffic for first one. So, group 224, dot five, do you have any information how to reach that particular source? Now the source, the receiver has to register themselves by using some Pin join messages.
Now, RP, now here RP is the root of the tree and it should know about both the source and the receivers. And it’s going to learn about the source by using some Pim register messages in the third step. And it’s going to learn about the receivers based on the Pim join messages. Now, all the routers in the reverse path here. Now, once the RP receives a join message from the particular router, Pim join messages. Now, all the routers in the reverse path towards the RP are going to install something called asterisk Star kama G entries and forward the join join messages hop by hop towards RP. Now, which means it goes hop by hop. Let’s say this router is going to send, I have a receiver here, it’s going to send an entry and they are going to join Star kamaji entries and then Star kamai entries towards the RP here.
Now, at this point, all the downstream routers towards the receiver from the RP, they will have S comma G entries. Now from the RP here towards the downstream router, they all have a Star commaji entries. And from the RP towards the source, it will have something called Escomma G entries. Now, there’s a fourth step. Now, at this point of time, now you have from sender will have S comma g entry, there’s a first point and the RP will have an eskoma g entries and all the other remaining routers. All the remaining routers, they have just an entry of star kamaji. So they just have that information that if they want to go to any multicast group address, they will send it to RP. They just have upstream towards RP. So RP knows exactly where is the source for that multicast group.
Now, the next step is it’s going to build a shade tree from sender to receiver through RP. Now, once the RP knows about the sender, the RP knows about the sender based on the pim register messages. Now, the RP also knows about the receiver based on the PM join messages. Now, the RP is going to send pim join message up to the source. Now, it’s going to receive the PM join messages and then RP is going to send that I want to join to this particular group. So that’s what it’s going to send. Now, all the routers in the reverse path again from the RP to the source. Install a stick comedy entries here again with output interface facing towards the RP again. And the multicast traffic begins to flow via RP to the receiver based on the star commercial entries.
Now, once it’s going to build a shade tree from the source towards the RP, you have a star commercial entries. So the multicast traffic goes through RP and from the RP it will flow back to the receiver with the star commercial entries. Now, this is the next step. It’s going to build a shade tree from sender to receiver via RP. Now, after that, finally the last step, it is joining the shortest path. Now, once the receiver here, now there’s a receiver here and we have a sender here, multicast traffic. Now, when a receiver sees the traffic star comma g here, that is a group address from the actual source. Now, actual source is here. Ten one from the actual source, they will initiate a pim join message towards the source.
Now, it is going to send a pim join message towards the source. It’s going to send the joint message directly to the source, ignoring the RP here. And it’s going to build the shortest part three based on S comma g entries again from source to receiver. Now, the routers will know about the actual source. Initially, they are going via RP based on this tarkamajan trees. Now, they are going to build a shortest path, path from source to receiver directly ignoring the RP here. Now, this new tree follows the optimal path from the source destination and it removes the RP for forwarding all the multicast property. It’s going to remove the RP, the need for the RP, because the source and the receiver, they know each other directly based on the shortest path.
Now, this process of switching from Asterisk kamaji, the star kama G to S kama G entries, we call it as a multicast shortest path tree switch over and once it shift to the shortest path, now the receiver router is going to send out a prune message to the RP saying that, hello RP. Now I got the best path to reach the particular source it’s going to inform to the RP, saying that it tells the RP that to remove this interface from the list so that it doesn’t need any multicast traffic yrp, because it has found the shortest path here. So this process switch over from acting well, source tree to the shade tree here. So initially if you’re running the spark mode, it’s going to run a source tree where your traffic will go via RP, but the receiver is going to figure out the shortest path from the source receiver and based on that is going to build the shade tree again.
2. Sparse Mode – Configuration
Now in this video we’ll try to verify the ipp’s passport configuration. It also will verify with some specific Show commands. Now, the first thing, I’m using the same topology of four routers here router one, router two routers, router four. And the first step I have already IGP configured with Osp of area zero. Now, that’s the first thing we are going to do. Anyway, IGP is something I pre configured just now. And if you verify schwa IP osp of neighbor, I should see router two is forming the neighborship with router one and router three. And then same thing if I check on the router one, if I use Show IP osp router four, I can see three neighbors. And if you verify the routing table, I should see the routes here, all the routes as per the diagram here. Now there’s a basic setup which we need to do.
And then the second step I’m going to enable IP pims pass mode on all the connect interfaces. And after that we are going to configure the rp. And then we’ll test by simulating a host or a receiver by using igmp join group commands. So let’s go to the second step configuration. So I’m going to enable IP multicasting on the routers. I don’t have multicasting enabled. I removed all the all the previous lab configurations here and S one by zero interface. I’m going to enable sparse mode here and same way on interface S one by one and the same way on the lan interface, sparse mode. I got the same configurations on the router two also. So simply I can copy paste these configurations. I can see the Pim neighborship establishes, establishes it just shows you the message.
And the same thing on the router three also. And because I got the same interfaces on all the routers on the router four also, I got the same interfaces. But on the router four I got one interface extra, that is S one by two which is connecting to router phi. And on the router phi I’m going to enable on F zero by zero. In fact, on S one by one, there is no S one by one interface here. So I can simply disable on that interface. If you verify show IP pim interfaces. Now, on the router file I enable on two interfaces. And if you check on the router four, if I use Show IP pim enables, I can see on the router four I have three Pim neighbors, three, three, one, that is router three, router One and router five. And if you verify Show Ippim interfaces and you can see on the router four, I have four interfaces enabled.
Now, the configuration is more similar to the thing what we did in the tense mode. The only difference is instead of enabling with Ipm tense mode, we are going to run sparse mode. Now, in case of sparse mode, the third step what I’m going to do is on the router four, we are going to simulate a host igmp join group. I have a receiver who wants to join to this group by using some igmp join message here. Now, even I explained you that on the router file we are going to simulate a host who wants to receive the multicast traffic for some group. So in my scenario, I’m going to say 224 five or five. Now, this way we are simulating a host in the land who wants to this is the multicast traffic. So once you configure this command, it’s going to send igmp join message.
But here where it is going to send, because in case of Spartz mode, it has to send the join message to the rp, right? But we did not configure any rp. Now, if you try to verify from any one of the router, let’s say on the router one, I have a multicast source sending the traffic and it’s going to generate some multicast traffic for this group. You can see I’m not able to get the reply here. Now, the reason is, if you are enabling with sparse mode, in case of sparse mode, it’s going to when you generate a ping message, it’s going to flood out of S comma g entries to all the routers. And if you don’t have any receiver, they are going to simply prune. But here you have a receiver. So it’s going to get the multicast traffic from here automatically. So it’s going to work based on flat and prune behavior.
That’s what happens in the dense mode. Actually not in the sparse mode, in the dense mode, but here, in case of sparse mode, we discuss that we must have an rp where the source must register themselves with rp saying that I’m the source. Let’s say the source is ten one and I’m the source who will be sending the multicast traffic for this particular group. And then wherever you have a receiver, receiver has to register with some pin join messages saying that I want to receive the multicast traffic for this group. And then the rp is going to send back a join message to the source with the star comma g entries. And then finally the receiver is going to figure out the shortest part for the source and it’s going to build the shortest part three scammergy.
And then it’s going to send out a prune message to the rp saying that I don’t want to receive the multicast traffic from the rp and it will request to remove the interface from the list. Now, this is a complete process. We have seen much more in detail in theory in the previous topics. Now here we are going to verify. So to make that possible, we need to configure an rp. Now, there are multiple ways to configure the rp here. Right now, I’m going to be the single step where on all the routers we need to simply say IP pimrp now, there is a command for that. We need to say rp address. The command is IP pimrp address. And then we need to configure any one router as an rp. Now, in my scenario, I have a loopback interface, loopback zero, that is 13 one.
And I want to make that particular interface as an rp. Now recommended to use loopbacks so that the loopbacks will not go down. They will be up most of the time unless you no shut down. Unless you shut down command, but not compulsory. You can make any physical interface also as an rp. Now, this is the static rp. We’ll talk about dynamic rp’s in the later on sessions, much more in detail in the advanced Pim concepts. But right now I’m going to configure the manual rp where I’m going to tell on each and every router saying that IP Pim rp address and the rp address is 31. Now, the same command we need to copy paste in all the routers. Now, let’s go to router Two as well. Copy paste the same command on the router Three as well on the Router Four as well as on the Router Five.
So we are just going with a basic normal rp configuration. Now, every router must know who is the rp so that they can send Pim register and Pim join messages. So we need to manually configure the rp on all the routers. There is an alternate way to learn the rp information dynamically also with auto rp or boots router concepts. We’ll be getting into that in the advanced Pin concepts. Now, once we configure this command, if I give show ipim rp mappings, I can see by default this is an rp address for all the group, for all class D addresses. Now, if I try to ping to the multicast group, I think I need to add one more command. Let me check. The thing is here it doesn’t work because the routers one, if you have a source here, they should know how to reach the rp based on the unicast message.
So, which means whatever the interface you are selecting as an rp, ensure that you are also advertising that interface in IGP also. And that is something we didn’t do. So if you don’t do that, then you will have some reachable tissues because it uses some unicast messages, pim register messages sent as a unicast the router one should know how to reach that 30 note network. There’s a reason we didn’t advertise the interface here. If you see here on the router Three, I have a loop back interface, but we did not advertise that. So I’ll go to Router Osp of one and I’m going to order as a loopback interface in the area zero. Now, once we advertise, we should have reachability to that particular appearance. Now, you can see I have reachability.
That’s the first thing. And one more thing you need to do is let’s go to router one and try to generate some traffic for 2424 five. Now, I can see I’m able to get the reply here and the reason is if you go to router one, router three, first we’ll verify on the router three, we’ll see Show ipp rp mappings and then the next thing I’ll say show Ipm root. Now, when I say show ipm root. Now this particular router rp, now it has S commaji entries for the source because now the source we have generated a ping request from the router one to this group and that is my host or the receiver. Now there are three interfaces on the router one. So that’s the reason you’ll see three entries. These are the three interfaces again. So it’s going to be the same.
Now it is incoming interface is S one by zero. It’s receiving the information from this interface. That’s the S one by zero interface and it is sending out of S one by one interface. So it’s receiving the multicast traffic on this interface, sending out of S one by one. So it’s receiving on this interface, sending back here. And there is a receiver here. And if you go back to router phi, if I use show ipm root for 224 five, the multicast group purchase which we are using here, I can see all the remaining routers also have an S comma g entries. Initially they will build Star kama g. Right now that’s what we discussed. Whenever you have a receiver you will have an asterisk comma g entries towards the rp and then the rp will have S comma g entries and then the source also will have S comma g entries. And once the receiver figure out the shortest part to reach the receiver, it’s going to build Eskoma geentries.
And then it is going to prune send a pruned message to the rp. That’s what we discussed. You can see it stops and you can see the rp information here. Now, that’s like a confirmation. Normally, if you want to see this, you can either use some demo of the debug commands in the back end to verify this process. But here I’m not getting into that. So now you can see without rp, if I don’t configure the rp in that case, probably you will not be able to reach the particular source to receiver. So they cannot reach because they must register with rp first based on the pim register messages and the receiver has to register with join messages. And then later on the receiver will build the shortest path to the source and it will send a prune message to the rp saying that it doesn’t want a multicast traffic from the rp.
Now, the implementation of the configuration goes very straightforward, unlike the way it works. You just need to enable the IP pimps passmode on the interface and make sure that your rp is reachable via IGP. And then to test out, we can configure a receiver to generate some igmp join messages by using some IP igmp join group command and then we can send a ping to that particular receiver. If you are able to ping, it confirms that simulating that there is a source here is able to send the multicast traffic for this particular group. Now, verification wise, it’s going to be the same like what we did in the dense mode. But the only difference is, in case of sparse mode, we must configure the rp. And the rp is the route for the for this tree. And it’s like a reference point. And you must confirm an rp. OK.
