And welcome to this CertificationKits CCNA video on Route Summarization. We’re going to go into further detail on Route Summarization that I might have mentioned in previous CCNA videos. How it helps reduce the routing table size, what auto summarization is by RIP and IGRP, what manual summarization with the null zero interface can do for us with manual summarization as well as some configuration commands.
The whole point of Route Summarization is to reduce the size of routing tables over here on this side of router we’ll call this router A. We’ve got the 10.100 subnet and what happens is they take this one subnet of 10.100 and break it up into smaller subnet. So we’ve got 10.1.1.0, 10.1.2.0, 10.1.3.0. So, Router A will know of all the individual subnets, but as far as router B goes; router B will simply have one routing entry into the routing table of 10.1.0.0. So what happens when any of these routers over here or any computer in one of these destination subnets on this side over here want to send the packets to let's say a computer with an IP address of 10.1.2.9; the routers will look at it and all of those routers will simply know hey anything with 10.1 in it send it this way. Anything with 10.1 in it send it over – over to router A. Then router A will pick up on the fact that there’s a two in there and go, oh that’s got to go out of this interface. So only router A will have detailed information where the 10.1.2.0 subnet is as far as all the other routers are concerned, it’s just simply the 10.1 network. As well as router B here, I’ve broken up a ton of subnets over here. We got 10.1.3.1 through 10.1.3.10 and I even broke the 10.1.3.10 subnet up into even more subnets using variable lengths subnet mask throughout all of the broadcast domains on this side of router B. So, here is 10.3.10.8, 10.16 and so on. So what could happen is there could be multiple Route Summarization going on over here if you wanted to. This Router would summarize all of the 10 3 that’s a three not a 10.10 0, subnets it basically summarize everything 10 3 10 0 until all these routers if you have anything that starts with a 10, a three and a 10 in the first three octet send it to me, he’ll worry about the dot 8, dot 16, 24 subnets.
Router B will send to router A, I’ll say hey anything with the 10 and the three in the first two octet, send it over to me. So as far as router A goes all of these subnets over here as well as the subnets of the 10 subnet, they will all just simply have one routing table entry in router A, Route Summarization. So, we’re going to take a look at how the different protocols go in and utilize this Route Summarization. All right let’s go over RIP version one, that’s the number there and IGRP remember there’s two versions of RIP, RIP version 1 and RIP version 2 big difference RIP version 2 sends the subnet mask along in the routing table update and supports variable links subnet masks. RIP version 1 does not, RIP version 1 and IGRP automatically summarize at network boundaries cannot turn auto summarization off and there’s problematic with discontiguous networks. Spell checker right there I had to get the spell checker for that one, so it's problematic with discontiguous networks. Let me bring up a diagram, so I can explain what I mean, here is the diagram, so I can explain what I mean by this discontiguous networks with RIP version 1 and IGRP.
Let's say over here, there’s like a 100 subnets coming off of this router and all this subnets are represented by the 10.1.0 subnet, so we’ve got 10 1 1 0 10 1 2 0 10 13 0 all the way up through 10 1 100 0. Way over here for some reason we got 10.2 network and everything starts with the 10.2 and then we got 10.2.1.0, 10.2.2.0, 10 – that’s supposed to be the 0.10.2.3.0 all the way up to 10.3 or 10.2.100.0. So, we’ve got a bunch of different subnets going on under 10.2 and 10.1 umbrella. Now, they are both a part of the 10. network. So, you might think logically the subnets here and down here would all be about the 10 but maybe they’re not, maybe they’re part of the 172.16 network so, we’ve got 172.16.1.0 subnet right here, 172.16.2.0 subnet down here, 172.16.3.0 subnet. So, the 10 network is separated by the 172.16 network. So, that is what discontiguous means, means it’s not all next to each other. So, there’s part of the 10 network go over here, part of the 10 network go over here, it’s separated by the C of the 172.16.0.0 network.
It wouldn’t be a problem with certain protocols like RIP version 2, the IGRP OSPF but with the RIP version 1 in IGRP it causes a problem. Let’s take a look at what that problem is. Let’s say we have a computer over here and most of the time it’s pretty happy guy. His IP address 172.16.3.1 and he wants to communicate with a computer over here with an IP address of 10.1.—I don’t know 9.8 so he’s on the 10 1 9 0 subnet and his IP address is 10 1 9 8. That’s this guy’s so they want to talk to each other, their buddies. The problem is router B is confused and the reason router B is confused is because his routing table is telling him different things. So, what happens is router B or when router A sends his information out to router B about his routing table all router A knows or all router A tells router B is hey, if you want to get since this is 172 subnet right here and when he sends the update out he says hey, I can get you to network 10. He doesn’t put 10 2 or anything like that he says hey if you want to get to network 10 come over to me, I’ll get you there and you’ll only be one hop away from network 10. So, router B has a serial zero interface and he thinks he can get to network 10 by going under the serial zero interface and it’s one hop away. Router C sends his routing table update to him and since this is a 172 network and this is called the boundary here as far as router C is concerned because he’s routing for a 172 and the 10 network he sends an update saying hey, if you want to get to network 10 this is B serial one you can get there by me. So, B will go OH, I can get to network 10 also through my serial one interface and you’ll only be one hop away. So, he thinks he’s one hop away from network 10 out of either interface. He doesn’t that there’s a 10.1 network and the 10.2 network because A and C since they’re running RIP version 1 or IGRP they summarised the 10 network and just send out to 10 no subnet attached with the class A subnet mask.
So, what router B is thinking, he is thinking he’s got two great path to get to the network 10 and they’re both one hop away. So, what he going to do is going to load balance. So, when this guy right here wants to talk to this computer over here router B left to its default configuration he’ll send some packets out to serial one and some packet out to serial zero. So, this guy is not going to get the whole message. Router A will get something destined for 10.1 whatever and that would be dumped on those packets just keep dumping those packets, never going to get there. So, some of the packets will make it, some won’t. If this path were to go down for some reason router B would send everything for the 10 network over here and not realising he can't get to the 10.1 subnet. He’ll just think it’s all right over there. So, they get confused because of this auto summarisation at network boundaries.
Here’s a little bit cleaner version, my previous slide I was looking at that, that’s pretty crazy looking there so I’m like clean it up a little bit. Here’s router B’s routing table just to reiterate, since networks or router C is that way, he considers in network boundary, he is between the ten and the 172.16 over here, router A is between the ten and the 172.16 and even though the subnet mask he’s using with this is 255.255.0.0. so, he knows the 10.2 is a subnet and he is using 255.255.0.0 when they send that update out to router B they send it like this 10000255.0.0.0 they summarize the entire ten with that and the same thing with A. 255.0.0.0 so, again in router B’s routing table, again there will be more there than just to ten’s they’ll have the 17216.networks is router B’s routing table. But as far as the tens go, he’ll think he’s got two paths, path one out of a serial interface one hop away and path two out of the serial one interface, one hop way. So, serial zero is path one, serial one path two and they are equal so, he will think he can load balance everything like that as far as he’s concerned there is no 10.1 and 10.2 separation router B is just going to think that there’s two ways to get to this network in their both equal so, he can load downs across to those.
If you are to turn load balancing off only one of these routes would be used maybe just this route so, all traffic going to the 10 anything what the 10 would go this direction even if it was for a machine like over here so, even if it was destine for a machine over here, if load balancing was off be able to just turn it send it this way doesn’t matter it would never get there. So again with RIP version 1 in IGRP summarizing at these network boundaries they do not support discontiguous networks meaning part of the 10 network over here and part of the 10 network over here separated by the 17216. RIP version and EIGRP do support this, we can turn off auto summarization or enable or disable at however we want, let’s take a look at that. So, like I was saying RIP version and EIGRP they send the subnet mask along with the routing updates, auto summarization can be enabled or disabled then we can also configure manuals summarization if we want to. So auto summarization and network boundaries can be enabled or disabled and we can configure manual summarization within the network if we want to. So, let’s go take a look at that let me bring up a slide, we’ll take a look at it and then we’ll go in configure it, I’ll show you the commands to configure it on a simulator.
Here’s that slide again, it shows 10.2 over here, 10.1 over here, 172.16 2 0, 3 0,1 0 and again router A and router C are where they considered network boundaries meaning between the ten and the 172.16. If we’re running a RIP version 2 or EIGRP we won’t have the same problem we run into with RIP version 1 or IGRP and the reason is we can turn off auto summarization we turn it off with the no auto summary command. So, I can turn it off right here and what I do is by going no auto summary is I tell router A hey when you send the update out, don’t summarize it so, when he sends it out 10.2.0.0 with a slash 16 subnet mask will show up in router B’s routing table he sends the full mask out of 255.255.0.0 with the 10 and the 2 in there. If I didn’t turn off auto summary then again this would be 255.0.0.0 and the 2 wouldn’t have an effect so it end up as a 10 in here instead of 10.2. So, router B will know hey, subnet 10.2 is out of my serial zero interface some more for using RIP it’s one hop a way if it’s IGRP it’s a cost of 80, 000 something. And he’ll know that subnet 10.1.0.0 is out of the serial one interface and one hop away if we’re using RIP version 1. Again the cost doesn’t matter, the key thing is, is that since the mask information is sent out with the update it’s not summarized, the routers can now tell the difference between the 10.2 network and 10.1 network, router B will have multiple entries in there but the subnets will be included as part of the entry. So, 10.2 and 10.1 instead of two entries for the 10.0.0.0 network it’s an individual entry for the 10.2 and the 10.1 network.
So, now and this guy right here he wants to talk to his buddy over in the 10 network somewhere, he sends it out router B goes oh I can’t get there through serial zero interface I have to send it out at the serial one interface if I want to get to the 10.1 network, and that’s what the no auto summary come in if we’d turn that off at the network boundary. Now within the network we can also do some summarization. Let me show you that no auto summary command first and then we’ll go take a look at what we would do within the network if you want to do some summarization. So I’ll bring up my simulator and the command that I would do if I had RIP running just like we showed in the slide I would go router RIP, and I would go no and if I forget how to spell it I can just use my question mark it’s no and this is the one down here auto summary. So auto summary turns it off and now RIP will send the mask information along with the update, so I could send information about the subnet instead of just the entire 10 network. Same thing, same command with EIGRP I just go router EIGRP here with the autonomous system number and go no auto summary, disables it at the network boundary.
I’ve brought up a different slide here and as we can see the 10.3 subnet is broken up into a bunch of smaller subnets we’ve got 10.2 right here in the middle, let me throw some zeros on there. And over here we’ve got 10.1.0 or 10.1.1.0 or 10.1.0.0 network with 10.1.1.0, 10.1.2.0, 10.1.3.0 these are not network boundaries right here this are simply subnet boundaries this whole environment right here that we’re looking at is 10.0.0.0 slash eight mask. I just happen to have chopped it up in two smaller subnets and I’ve even chopped the subnets up. There are multiple places that we could use a little bit of route summarization going on here, it’s not necessary for let’s say if there’s a router over here in the 10.1 subnet, it’s not necessary for that router to have all of these routes for the 10.3 subnets of – there might be a couple of 100 routes for the 10.3, 10.3.1.2.10 –10.3.200 or there might be 200 plus subnets over here. And since the network is designed appropriately it does not make any sense for let’s say router A to have every subnet in the 10.3 network, there should simply be one entry, 10.3.0.0 you have anything with the 10.3 in the first two octets send it over to router B, helps minimize the size of the routing tables.
Now, since these aren’t network boundaries auto summarization will not have an effect here, so if we want to reduce the size of the routing tables we can go in and summarize the routing tables ourselves with manual summarization. I could do it here because right here I’ve chopped the 10.3.10 subnet into a multiple subnets. I’ve got 10.3.10.8 subnet .16.24.32 and it keeps going 40, 48 I can keep doing that, so what I should do here is I should summarize and tell all the other routers hey if you want to get anything to the 10.3.10.0 subnet or network just send it over to me I’ll worry about the details. Here for route and that would go – that would cut down the size of the routing tables for all these other routers in here in the 10.3.0.0 network. Over here at router B we could summarize and tell all the routers hey if you have anything with a 10 and a 3 in the first octet cut that down, just send it over to me you don’t have to have 200 routing table entries. So we can cut it down it just to 10.3 and if anyone needs to get to a 10.3 whatever send it over to me and I worry about where it goes from there. Router A right here we’ve got 10.1.1.0, 10.1.2.0 there might be a 100 plus subnets over here. So what we should do at router A and say hey if there is anything with a 10.1 router A tell router B if there is anything with a 10.1 send it over to me, and I’ll deal about the – worry about the details.
Let’s take a look at how will would manually summarize that information. So with a cleaner slide again, I’m going to mark the spots where we’re going to do some summarization we are going to summarize here at router, let’s call this router – I don’t know D, we are going to summarize here at rather B and we will summarize here router A. Now router a we’re going to summarize everything that has a 10 and a one in the first two octets. So it’s going to be summarized with 10.1.0.0 even though there’s 10.1.1.0, 10.1.2.0, 10.1.3.0 subnet, all those subnet start with the 10 and a one so as far as the routing table for all these routers go we’re going to summarize it into a 10.1 entry. Router B is going to summarize the information and that information is going to be summarized of 10.3.0.0 going out to router A so anything with the 10.3 router A is going to know to send it to router B. And within the subnets over here at router D we’re going to summarize the 10.3.10 subnet. So there it’s going to be summarized and router D is going to tell all the routers over here hey anything with the 10.3 and a 10.1 three octets, so it will be slash 24 mask send it to me, so he’ll send that information out to all these routers.
So there’s a command that we would use to that and we’ll go into the CCNA sim and look at but first we are just going to do it in the diagram here. So let’s say this is router A’s serial zero interface, this is router B’s serial one interface and this interface right here is router D’s serial one interface. So on this interfaces we would go in and use this particular command, I’m going to type this command out because of my handwriting it’s a little bit sloppy. Here’s the command typed out on router A I would go to interface serial zero and again RIP or EIGRP would be running and I was just type an IP summary address and then the protocol, I would just type in RIP if I was running RIP or EIGRP and the autonomous system number which might be one or whatever but I would type in the autonomous system number here. Then I just type in my summarization so here I am trying to summarize 10.1.1.0, 10.1.2.0, 10.1.3.0 into one routing table entry so I would type 10.1.0.0255 25500 so what this mask does is it summarizes all of those it cuts it back and says only this part is part of the network address same thing for router B. IP summary address the protocol 10.3.0.025126.96.36.199 this one will be a little bit trickier because I would go IP summary address the protocol and it would be 10.3.10.0 and my subnet mask will be 255.255.255.0 because I’m trying to summarize the entire 10.3, 10.0 subnet into one entry. The normal subnet mask is dot here is 255255255.248 so I’m using that and I dial that back to just this to lump all of the subnets to 816, 24 and 32 subnet into one routing table entry.
So that’s auto or manual summarization and again auto summarization we’ll use the no auto summary command to disable it at a network boundary. If we want to do some manual summarization here we got to remember that these are not network boundaries. This is just simply summarizing within the ten network this whole environment is part of network 10.0.0.0. Let’s go into the simulator and just look at that auto summary command just so you can see it on the command line and then we’ll review what we’ve gone over. I’m bringing up my simulator and I’m going to go in and go to interface serial zero just like we saw on the slide, this is palaestra one and what I do is I go IP and if I forget I can just use my question mark, summary address. So, summary address and then it wants the protocol, know this I only have two options RIP and this is RIP version 2 by the way or EIGRP. So if I did EIGRP I'd have to specify the autonomous system number.
RIP and then it wants to know what the address is, so according to our slide I would put 10.1.0.0 if it was on a router A in the slide, this is palaestra one, 255.255.0.0 by doing a question mark it simply says carriagereturn meaning hit enter and you’re done. So it’s a fairly easy command, the main thing is just making sure you’re IP is planned out appropriately in your network. So, here again we've talked about RIP version 1, RIP – and IGRP, auto summarization at network boundaries again can be problem there can be problems with discontiguous networks. RIP version 2 if we do have discontiguous networks or EIGRP would be a better option can be enabled and disabled at network boundaries auto summarization can, and manual summarization can be configured within the network if we want to cut down on the size of our routing tables. OSPF we don’t have to sweat because our area of boundaries it’s going to summarize automatically and again when we design OSPF or IP design must coincide with the areas to make it work efficiently and if we do that right, we don’t have to worry about it.
I have brought up my simulator and I've gone in and configured it with summarization so if we do a show run, we can see that at interface serial zero with an IP address of 10.2.0.1 I've configured a summary address, so I've done summarization. IP summary address again RIP 10.1.0.0, 255.255.0.0 on Ethernet 1 I have an IP address of 10.1.1.1 with the 255.255.255.0 mask notice the masks. This one is a lot longer its class C masks or slash 24 mask as opposed to a slash 16 mask with summary address. And that’s because theoretically off of my Ethernet zero interface I might have more routers coming off that interface with a bunch of more subnets that are all a part of the 10.1 network, I have got maybe a 10.1.2 subnet 10.1.3 subnet all beyond the Ethernet interface. So I've summarized the whole 10.1 subnet with this command right here. If we look with the RIP configuration router RIP I've turned on version 2 so it actually works, specified I’m routing for networks 10.0.0.0. What I want to show you is in the routing table, show IP route, this extra route gets added in here 10.1.0.016 is a summary and it says null zero. This null zero interface is something I want to address so we have to be able to see it, to be able to talk about it. What it is, is if RIP knows where the destination subnet is, it will go ahead and send it on its way so if it knows where 10.1.2.0 it will go ahead and send it on its way out of its Ethernet zero interface because that’s where it would be in existence.
The null zero interface is there in case this router doesn’t know where the destination subnet is, it’s basically like the garbage can it routes the packet to the null zero interface and it just floats off into space somewhere never getting to the destination, so it’s kind of like it’s garbage can jus tosses in the trash if it doesn’t know where the destination subnet is. That’s the whole point of summarization is, the next top router will know that everything with the 10.1.0.0 is on this router or beyond this router. So it doesn’t have to worry about knowing where 10.1.10 is everything gets sent here. This is the only router that should know the details of the 10.1.0.0 network as well as maybe other routers within it. So null zero interface is there in case this router doesn’t know where maybe 10.1.101 is the subnet 10.1.101 it will go ahead and dump it into the trash and that’s a null zero interface, it’s kind of like the trash can. So we have talked about route summarization, what summarization is, how it reduces the routing table size, auto summarization, manual summarization within the network, the null zero interface, again kind of like the trash can for the routers and configuration commands we've used to go over it. I hope you have enjoyed this CertificationKits CCNA training video with Route Summarization.