3. OSPF Part I Introduction Neighbors

hello and welcome to this tutorial on OSPF this is the first of several where we talk about the theory and concepts behind OSPF after this then we'll move into the configuration tutorials so here in part one we're going to introduce you to OSPF and we'll begin by looking at the basic characteristics a lot of just useful background information about OSPF and then we're going to jump into neighbors and how they're formed there are different types of neighbors that can be formed in OSPF there are different stages of becoming a neighbor and there are some really important details that we have to cover so that you when you're troubleshooting it or when you're thinking about design are really well-informed okay so let's get started with the overview if you've already watched the link state routing protocols tutorial then you know that OSPF is a link state routing protocol so it has the characteristics of creating neighbors and exchanging network information by using LSAs it also keeps a local link state database where it has all the information about the network and it applies an algorithm to that data in order to determine the best routes through the network now OSPF has two versions 2 and 3 3 is the most recent has some enhancements over version 2 including the support for IP version 6 so if you haven't yet it's likely you soon will have exposure to version 3 at the workplace as more and more people adopt IP version 6 OSPF being a link state routing protocol offers much better convergence times than your standard distance-vector protocols so OSPF quite simply is better than rip it's a lot more complex but it's also a lot more robust so overall we get a lot more benefits from OSPF in terms of performance and that's what we want if you look at OSPF it really doesn't require much bandwidth in order to operate so when neighbors are talking to each other and exchanging that OSPF information not a lot of bandwidth is needed and that's a good thing let's save our bandwidth for actual production traffic now OSPF supports a lot of the features that the modern network requires like Sider support variable length subnet masking and authentication so that's one reason why it's such a popular choice in addition to the fact that it has such high convergence times and it's an open standard now when you talk about OSPF the metric involved is known as cost and this cost is derived from the bandwidth of the links on your network so quite simply the higher the bandwidth the lower the cost and OSPF is just like us we want to pay as little as possible so it likes the lowest cost to get to a destination there is a design approach to keep in mind when you implement OSPF on small networks there isn't so much of a concern but on large networks it's really a requirement to put a lot of thought and consideration into how your network will use OSPF and so in the design tutorial you're going to come across something called areas and that's just a way to kind of segment your network by using OSPF and the area's perform differently they have different names to them also related to this design approach is that routers when they use OSPF they can take on different roles and we'll cover that as well basically one router running OSPF can have a different function than another router and these relate back to the areas and the overall design considerations okay so these are the the basics the the the basic characteristics of OSPF commit these to memory now let's go ahead and jump into how OSPF actually functions now like most routing protocols OSPF uses the concept of neighbor and basically all that means is another device that's we're also running OSPF and that device i can exchange information with about the network now neighbors are created in OSPF by using something called a hello packet and this hello packet contains a lot of OSPF information about each router and these hello packets are exchanged between all routers running OSPF if router a were to send out a hello packet it would use a destination address of to 2400 five that's a multicast address reserved for OSPF that means any other router that is also running OSPF is going to accept traffic sent to that multicast IP address okay so an OSPF router doesn't necessarily know who or where its neighbors are so neighbors are discovered dynamically using this hello packet and by sending it to the multicast address inside this hello packet is a lot of information you have something called the router ID this is a 32-bit value and it is the OSPF ID of a router so it identifies each router in the OSPF world also there's something called an area ID we mentioned the hierarchical approach to implementing OSPF and when we we talked about areas being a way to segment your network well each area has its own ID and so that information would be in there as well there's also other stuff like a router priority something called the hello interval and a dead interval it sounds kind of scary and then stuff like a designated in a backup router and a list of all of the other OSPF neighbors a router has made don't worry we're going to go through all this information so by the time you're done with all the OSPF tutorials you'll know what these mean now the hello information has to match between two different routers in order for them to become neighbors just because you receive a hello packet doesn't mean you're necessarily going to automatically become a neighbor with the other router so the subnet number and the subnet mask of a particular router that has to match the subnet and subnet mask of another router so routers a B and C should all have the same subnet number and subnet mask also the hello and the dead intervals have to match between two routers and the OSPF area ID has to match and there's even more requirements if you're using authentication you obviously have to be able to authenticate in order to become a neighbor and there's something known as a stub area flag that too is outside of the realm of the CCNA study but we're mentioning it here so that you have a complete picture of what's involved but just put that one on hold for now we don't cover that the hollow interval is quite simply the timer for sending out hello messages the dead interval is related to that and this is the time a router is going to wait to receive a hello before declaring a neighbor's down so the hello interval would tell router a for instance to send a hello message every X number of seconds and the dead interval would tell router be for example to wait a certain amount of time in order to receive a hello message from router a and if it doesn't receive it then there might be a problem and we need to mark router a as down potentially so let's take a look at the hello process in action and it's relatively simple it starts off with a router sending out a hello packet to the multicast address so let's say router a sends it out on the link between it and router B now router B is going to receive that packet and it'll start to examine all of the information we dis mentioned in the hello packet now if all of that information checks out in other words everything that should match matches and there are no problems router B is going to send a hello packet back to router a however when it does it's going to include a little bit of information about router a specifically it's going to list router A's router ID inside that hello packet now this may seem kind of strange but watch what happens when radder a receives that hello packet back from router B it's going to see among other things its own router ID inside that message and that's actually a key bit of information because as once it sees that router ID in that packet it knows that router B receive that hello message and it has started to communicate with router a okay so as the hello process continues routers are going to move into different states of becoming neighbors and each state kind of signifies a different level of progress in becoming neighbors so if two routers used to be neighbors and for some reason they stopped being neighbors perhaps a link between them went down or there's a configuration change the neighbor state is called down so they used to be neighbors but they just can't communicate now however the next step in becoming a neighbor is known as the innate state and that that's just an abbreviation for initializing initializing their neighbors are starting to talk to each other so I'm starting to send out my hello messages okay neighbors are initializing that process now if the hello information is okay like router B thought it was and then it's sent back that hello message with router A's router ID in it and once that process happens for router B as well in other words router a sends a hello message back to router be with router B's router ID in it that way router B knows that router a got it then those two routers move into what is known as the two-way state and this is a state that really just means that a bi-directional communication has been established between router a and B now at this point the routers are ready to go into one of two directions they can decide to exchange topology topology information the the link-state advertisements and if they go through that process successfully they'll eventually become full neighbors or fully adjacent to each other that's one option they have the other option is they cannot become full neighbors and in fact they're going to elect a different router on the network to which all routers will become fully adjacent and they'll leave other neighbors in the two-way State now that might seem kind of confusing so let's dig into that a little bit more now the state of a neighbor depends on how the link to that neighbor is defined and in OSPF this definition is known as a network type so one network type is known as the broadcast in other words broad broadcast frames are supported like on Ethernet and many devices are connected to the same segment another network type is known as point-to-point and here only two devices are connected to a single Network segment so in this diagram here we only have point-to-point network types between the different routers now each of these network types has a default behavior so on a point-to-point network type the router is by default will proceed to exchange their link state databases and eventually become fully adjacent to each other so here router a and B they have a point-to-point network segment between them at the point where they become two-way that we just talked about that two-way neighbor state their default behavior would would dictate that yes let's go ahead and and begin to exchange information and eventually we'll become fully adjacent however if your network type is not point-to-point but instead it is broadcast in other words a lot of different devices are connected to it at the same time then the default behavior actually changes so here we have a shared network routers a b and c all connected to the same network segment and here the default behavior is do not exchange your link state database by default in fact do not become fully adjacent to all of the other members on this particular network segment by default rather what they want to do is in order to minimize the amount of information being being exchanged between all of the routers connected to this one segment they're going to go ahead and elect a focal point or a center of the network and each router is going to become fully adjacent only to that center point they will not become adjacent to each other now this center point is called the designated router and so what would happen here is routers a and C let's say would elect router B to be the designated router that would mean that router a becomes fully adjacent with router B and router C becomes fully adjacent with rather be however routers a and C do not become fully adjacent rather they will remain in that two-way state and it is up to router B to relay the information between router a and C so that their link state databases are completely the same exactly identical to each other so what we're doing here is if we had 10 more routers on this network segment there would be a lot of traffic going on between all the routers to establish neighbors and to exchange information all the time it's not really efficient to do that on a shared network segment so that's why we elect this focal point the designated router now there's also something called a BD R or a backup designated router and this is quite simply a backup to the primary designated router should it fail everyone else can still remain adjacent to the backup designated router now we mentioned that the designated router would be elected by router a and C for instance well this election is based on the router priority and that was one of the fields that we listed in the hello message the router priority well this is a value and it's a value between 1 and 255 and each router can have one assigned to it and when they all compare their router priorities the one with the highest value wins if they're tied for instance then they resort to the router ID to determine which one would then become the designated router again that's the OSPF router ID that 32 bit value oftentimes the router with the second highest priority would then become your backup designated router if you don't want a particular router to have a chance to even become a designated router it doesn't even want to participate in the election you can give it a priority of zero and that way it has no chance of becoming either a designated or a backup designated router and so each one of these routers would have a value a priority value to it and here we said router B was our elected designated router and here you can see why it has the highest router priority and so a would become adjacent to router B and C would become adjacent to router B as well okay to summarize what we covered there are some basic characteristics to OSPF it's a link state routing protocol there's version 2 and 3 it supports Sider variable length subnet masking authentication the metric is known as cost it's based on bandwidth review the beginning introduction to this tutorial and make sure you're familiar with all of those characteristics of OSPF then we talked about neighbors and we talked about how hello messages are sent what's inside of them and the overall process of how two routers get to the two-way state of becoming neighbors we also mentioned that the hello message is sent to the multicast IP address of 224 0 0 5 remember that's a dedicated IP just for OSPF make sure you remember that one then we talked a little bit about the different states that two routers could be in when they're becoming neighbors anything from being completely down to initializing to the two-way state and then we didn't talk about these next two exchange start and exchange they're a bit outside of the scope of this tutorial but we're going to touch upon them in the next tutorial on OSPF and then finally we touch briefly upon the full adjacency and again we'll get into more details in the next tutorial on that state as well be familiar with this order and which eats what each state actually means and then finally we talked a little a little bit about designated routers backup designated routers the fact that they're elected on the network and the router priority and the router OSPF ID come into play when an election is made okay so that's it that is the first tutorial on OSPF thanks for watching

Channel: System Engineer

Views: 2,624

Rating: 5 out of 5

Keywords:

Id: 6AIslkVX06Q

Length: 18min 2sec (1082 seconds)

Published: Tue Apr 25 2017