100 IPExpert LSA Types 1 2 3 and Virtual Links

[Music] the next thing that we are going to go over is going to be the different OSPF LSA types now before I can take before I can talk about the LSA types I believe I owe you an explanation of what actually link state is so we call OSPF a link state protocol but what is the link state what do we mean when we say the link state well the link state is really a couple of things one is it's the description of the link in terms of cost and type so here we have some metric that is associated or type and as I said in the database this can be described as stub link point-to-point link or transit make sure there are only three network types that are inside the database and also another thing that link state actually is is relationships with other routers on the link so when we say the link state this is what we mean this is the information that OSPF routers will have in their databases about every single link every single interface in the network in their area so when we talk about different LSA types this is what we have to keep in mind what part of this will be contained in that LSA type so let's start with the easy ones so we have type number one which is also called the router LSA now router LSA is generated by all OSPF routers and the data contained there the important pieces of information are the router IDs router link States I'm going to say router links so this is the important information that is contained inside this LSA now this field here is incredibly important what is the flooding scope of type 1 LSA now what do I mean by flooding scope so funny scope is how far in the network can our LSA travel unchanged so when I when I have for example routers that are interconnected like this how far in our network will this LSA let's say generated by this router here so there is some LSA type one actually there is LS a type one generated by this router one how far in this network will it travel now that depends on how many areas we have if this was one area for example the flooding scope of this LSA would be this so this would be one area but let's say that this here is another area I may not be doing this correctly so let me just dumb so let's say that this here is another area now LSA generated by this router here will have this as the flooding scope and let's say that we have another area that actually sits here and the so this is LSA type one also I do apologize this is LSA one but from router 2 this is router 3 so the flooding scope here for this LSA would be this area now if this network was just a single area if this was one area the flooding scope of type 1 LSI's would be this whole thing so this is what I mean by flooding scope the flooding scope for type 1 LS s is a single area and let me clarify one more thing what I mean by flooding scope is when this router generates the LSA it inserts it in its local database this means that this LSA identical must exist on this router and it must exist in its identical form on this router because inside this area our databases must be identical so none of these routers can actually change any information from that LSA they can only send it as is and this sending as is is flooding so I'm just going to send as is this is what I mean by following how far the network will this LS a travel as is so to give you a practical example and this is the the network that I'm going to use to examine all of the LSA so I'm just going to draw four routers here and let's say the term this was our one this was our two this was our three this was our four let's say this is area 12 this is area 0 and this is area 34 let's say that there is that r1 here generates type 1 LSA this is where this LSA is going to stop so it is never going to leave the area in which it was generated now our four here may be generating its own well not actually maybe it will be generating its type 1 and this is where this stops now r2 may be generating its own type 1 so this is where this information stops r3 is generating its own and this is where it stops now mind you inside area 12 r2 is generating type 1 LSI describing its own length in this area and so does r3 in this area here now important thing to note here is that on r2 there are two separate databases and so are on r3 so the database here for area 0 is separate from the database in area 12 and here for area 34 so this is the Green database this is the red database this is the red database and this is the blue database so these routers here have separate databases which means that these routers the border routers will actually generate to type one LSS 1 in area 12 in case of our two so this is our 2 and this is our 3 so this one here will be generating type 1 LSA in area 12 and in area 0 and R 3 will be generating type 1 LSA in area 0 and in area 34 area routers like for example r1 and r4 are going to be generating only the LSS for their own links in in this area also our 2 here will be generating type 1 LSAs only for the links that exist in area 12 those links that are in area 0 will not be sent in the blue domain and these blue links will not be advertised in type 1 LSA in red domain now one more thing that might be misleading here is router will only generate 1 type 1 LSA per area type 1 LSA contains or may contain I should I should say information about multiple links so for example our two here may have ten links in area 12 but it will still generate just a single type one LSA and in this single type one LSA we are going to have ten link entries describing those ten links in terms of cost and their relationship with other routers same thing for area 0 and same thing for area 30 for going back to our table here let's take a look at another LSA type this is LSA type 2 this one is called the network LSA and it's generated by dr routers now dr routers as you remember have two functions the primary function is to create that virtual hub-and-spoke for exchanging the information on the local segment and the secondary function is to let everyone else in the network know about this arrangement and this is exactly this bit type 2 LSA is that secondary function so dr routers on the segments are actually going to inject their information about the netmask and connected router IDs so if we had a network that looks like this so r1 r2 r3 when let's say that this is our designated router so here is where we have our adjacencies now r1 and r2 and r3 are both going to generate their type 1 LSA and in this type 1 LSA they're going to describe their local interfaces here so the description on let's say r1 may say this is let's say cost is let's say 100 and the designated router on the segment and let's say that the IP address here is 102 168 1 2 3 0 / 24 dot 1.3 not 2 and here we're going to have the information that the designated router is 102 168 1 2 3 3 the description of the intro so this is in type 1 LS a type 1 LSA on r2 will have let's say that the cost is 50 so we the cost doesn't have to be the same right this could be a gigabit into this could be fast ethernet interface this could be a gig interface so of course sometimes the costs might be different and we're going to have the information about the DR that is 192 168 1 2 3 3 this type 1 here is going to contain the information that let's say cost is let's say also 100 and that the DR is 192 168 1 2 3 3 now the important thing to note here is that every link state every LSA in the database has an identifier now the identifier for type 1 LSA is actually going to be the router ID so this is going to be r3 sorry this is going to be R 1 and here we're going to have that link state ID here is our to whatever that value might be now because this router is the dr it is actually going to generate a type to LSA as well now the idea of this LSA is actually going to be 192 168 1 2 3 3 now take a look at this all of these routers are actually listing the IP address of the designated router as the key to actually look up type to LSA so they are telling other routers if you need more information look into this LSA so the ID of type 2 LSA will actually be the IP address of the designated router on the segment information that is going to be contained in there is going to be the netmask and connected routers so in this case we are going to have the list of r1 r2 and r3 now these are the router IDs so whatever those values were but the important bit here is the netmask because inside these type 1 LS s there will be no information about the netmask because this is the link type of transit when we have the transit link you are not specifying the net mask the net mask needs to be looked up inside type 2 LSA that is going to be generated by the dr so again link type here is transit and link type here is transit so going back to our table the information contained inside there is going to be the net mask and the connected router IDs the flooding scope just like with type 1 LS s is the area because this information makes no sense outside the area so technically speaking if you remember my description of what LSA is and it's the description of the link wheel in terms of cost and type and the relationship with other we can see here that for the shared segment like this the information about links is delivered individually by all the routers but the actual relationships on the link are delivered by the dr only so this is why this is how we are actually optimizing the information exchange because there is no point in this information being sent by all three routers or 50 if we had 50 on the segment so again you see how the designers of OSPF actually thought this true why this information here helps us out so going back to our table the next information the next LSA type that we have is type 3 which is called summary LSA and this is where some of the major confusions in OSPF begin so again I'm going to use the example of this network here which is a perfect example for what I need I can just tap delete this a little bit okay now let's say that we have a network n that exists here so this network n lives in area 12 it will be advertised in area 12 as type 1 LSA and optionally type 2 LSA so type 2 may or may not exist depending on what link type this was if it was a point-to-point or stop we won't have it if it was broadcast or non-broadcast will have type 2 so this information ends here now based on this information here r2 is going to calculate what is the cost to reach n now mind you this was a simple network here maybe we had multiple devices here in between right so we don't know what's actually sitting in this network here and it's borderline irrelevant because what's going to happen r2 is here based on this information here and based on the costs of of this link here in this link here and whatever the path it needs to take it's actually going to calculate the cost to reach n so it's going to take the information from type 1 type 2 other information from the database it's going to compute it and it's going to generate a type 3 LSA basically summarizing the LSA information summarizing the database information and it's going to inform the routers in area 0 that there is some network M reachable through r2 with cost X the calculated cost X ok so this is the what we calculated and this is the value that we are going to insert here so basically we are saying this network this cost to reach it now this is what I was mentioning that this is a very very distance vector like behavior we have the route in a routing table this is our cost to reach it we are going to advertise this to our neighbors now important thing to notice here is that the originating router of this information here is no longer r1 it is our - so r2 generates this information now r2 will be identified as the originating router for this LSA inside the database and this originating router is going to be the router ID of r2 now r3 knows about r2 because it also had type 1 LSA describing are two links now what about our four can our four reach the network and advertise by r1 well if r3 was to forward this information out to flood it out you will have information here that this is reachable through r2 and r4 has no idea who our two is because this information stops here this is why the flooding scope of type 3 LSA is actually a single area so this type 3 LSA stops here but r3 is actually going to generate its own type 3 LSA so it's going to look ok I have Network n so it's going to take this information here and it's going to calculate its own table cost to reach n is cost to r2 + X make sense and let's say that this is y so this is the information that we have so r3 is now going to generate a new type 3 LSA and it's basically going to say that n is reachable through r3 with cost why ok as you can see we are taking the route or information advertised by our neighbor we are adding our own cost and we are advertising this information out very very distance vector like behavior now when we have distance vector behavior we always have problems with loops so how do routers prevent loops in distance vector environments well they use things like split horizon they lose things like count to infinity etc so let's take a look at how OSPF prevents loops in a multi area scenario now this is going to get very involved and if you are hungry please don't look at the screen now because I'm going to start drawing some cookies so let's say that I have three OSPF areas or three Network areas now these three Network areas and of course every time I need them to be separate so apologies for that I'm just going to do this one more time so I have one area here and let's say that inside this network I have some relationships like this so I'm going to take this and just clone it and clone this one as well so there are two ways these areas can be interconnected now there is a one way that is used by OSPF and there is one way that is not used by OSPF the way that is not used by OSPF would be this in this scenario here the routers are always members of areas and links are the borders between areas this is how is is works but you don't care about is is what you care about is OSPF so the way OSPF works is that links are members of areas but routers can actually be members of multiple areas so what we need here are the routers that are going to be sitting between the areas that have links that are members of individual areas but they themselves may belong to multiple areas so these routers here are called border routers and border routers have multiple databases so let's say that this here is area one this here is area 2 and this here is area 3 now even though I can't hear the sound from you I see at least one or two you screaming at me hey but this is not how it works you can't do this in OSPF and my answer to that is yes I know but bear with me there is a reason why I'm doing it this way so let's say that here inside area one we have some network in and we want to advertise this network to the rest of the network so let's see what happens there this route gets advertised inside this area let's type one LSA so this is an internal route inside this area so remember these routers here have two databases now these border routers and what what I'm going to do here I'm just going to observe information going this way of course the information might be going in the opposite direction as well but they're with me so let's say that this here is border router 1 this is border router 2 and this is border router 3 so border router 1 now has this information in its database and it's going to create a summarized information and it's going to inject it into this area here as type 3 LSA so this information as we know is going to flow here and this must flow to all these 3 routers and this router here because the database must be identical now this router here is going to create its own summary of this information and it's going to inject it into this area here so when this information gets injected into this area here we create that information in the database here so now this router 3 has this information in its database and depending on the cost of these links here it might actually inject it back into this area so we get this information fed back and fed back here which means that we need to inject it here and we are running the danger of actually creating a loop in our network now this would be the problem if there were no special safeguards against this in OSPF but fortunately for us there are actually safeguards to prevent this from happening the first safeguard that we have and let me go back just one page before because I want the clean copy of this diagram the safeguards are that not all the routes in OSPF are the same so the first thing that I'm going to say here is not all the routes are the same OSPF divides routes in three categories internal inter area and external rats now internal routes are those routes that are carried in type 1 and type 2 LS A's these are typed through L type 3 LS A's and these are type 5 and type 7 LS s I will talk about them in a moment as I said sometimes chicken in the egg now how does this help with the problem that we had before well the way it helps is that OSPF implements one of two possible routing philosophies so let me just create a new slide here and there are two routing philosophies as yours truly is concerned one is so-called hamster routing and mind you these are my names and the other one is hot potato road now what I do I call it comes to rocket maybe it's not thus comes throughout may be the more correct example would be squirrel routing but I have this mental image I don't know if you have seen Disney's cartoon called the Ice Age and you know that tiny little squirrel that has that blinky eye and has that not that it carries around the whole movie and every time he sees that that he goes fine and keeps it now that's what comes through outing is try to keep traffic as local as possible try to keep the traffic local so OSPF is going to try to keep traffic inside an area as much as it can now as opposed to that hot potato routing is for example so this is done by OSPF hot potato just to give you an idea what hot potato routing is is get rid of it as soon as you can and this is what for example BGP does now if you remember what you haven't talked about BGP yet but you probably know a little bit about it BGP prefers external routes over internal routes now the whole idea here is that if I have an external route and I have an internal route assumption here is that this traffic at some point needs to leave my control now I want to get rid of it as soon as I can but why would I waste my own bandwidth why would I waste my own resources to send this traffic to an internal neighbor just to send it out of my network when I can get rid of it right here right now so that's hot potato routing so OSPF actually implements what i like to call the hamster routing it tries to keep traffic as local as possible so how does that help with this here well if we take a look at this this is the order of preference given two routes so this is preference internal route so if i have inside my network an internal rat i'm always going to inside an area give preference to keeping traffic local then sending it out so this is going to be preference number one keep the traffic local and only if i don't have the internal route only then am i going to send it out of my area and only if i don't have that I will send the traffic outside of my autonomous system outside not autonomous system out my OSPF domain so the route preference here helps just a little bit how does it help so going back to this complicated example that we had here so what we clone this one if you take a look at this example here the reason why this situation here happened why this information was injected and the resulting traffic that might have actually followed this way and traffic looping is never going to happen because inside this area inside this area one we are always going to prefer these internal paths no matter what the cost says outside because if we take a look at this network from the perspective of area two routers or three of area one routers this is how this network looks oops that's not what I wanted so I guessed that I need to get just a little bit creative so bear with me there so this is the view so this is our area one and the view of this network so we have this border router here let's take a look at how this router sees this network so this is the border and this is the border here so this router here knows about these borders but it doesn't know anything about the network there be dragons the only thing it knows is that some networks are reachable somehow behind these borders but it has a full visibility of inside the network and it's the case of the devil I know is the better than the devil I don't know so this is why OSPF tries to keep traffic locally here now even though the metric here could say hey look this metric is a little bit better would you like candy this is a problem right because we don't know what is happening here we don't know how true the information that is being fed here and injected by this ABR is but you know exactly how good information inside the area is so the oil space is going to keep traffic locally so this is going to prevent the actual traffic from looping so this will not happen we are never going to send the traffic outside the area only to receive it back but it doesn't help us with this problem here when we had the actual routing information being injected from one area into the area from where the traffic originated so this is where the second part of the solution comes in to pay play and the second part of the solution is of course something that most of you have probably been screaming about and this is that this is not exactly how SPF works so my set up of these three areas when they are interconnected as they are is not possible oh SPF does not allow for that topology to be built what OSPF requires is a dedicated backbone area that all other areas connect to so this is a valid OSPF topology now the one that we had before is invalid so what we can have here is area one we can have here area two but this here needs to be the backbone area backbone area in OSPF of course is area zero technically speaking it's area 0.0.0.0 because area ids are also 32-bit numbers but in iOS we can just call it area zero basically what OSPF requires us to do here is to implement hub-and-spoke at the area level now if you have hub-and-spoke can you actually have a loop in hub-and-spoke environment if you don't have spoke to spoke connectivity you cannot this is loop free by a definition so this is why we have area 0 why we have the backbone area because now this internal information that we have here in these routers when it goes outside the area when it's injected here and when it's sent to this router it will be injected into this area here but there is no way that this information can find its way back there is no way it can go back because we have a loop free environment inherently loop free environment but there is of course a problem with that because this what I'm going to do next is also a perfectly legitimate OSPF topology so bear with me until I clean this up a little bit so now let's say that them are on our area to here instead of having just one border router we have more than one border so we are going to have let's say border 1 and border to this or let's make it even more fun like this so this is 1 this is 2 so now what we have here is a potential for the look so we had this internal information that was now injected into this area here and eventually both border 1 and border 2 in their green databases are going to learn this information so now let's say the border one inject this information and this information now reaches border - what are - now should take this information and actually inject it back into the database which means that eventually this information might find its way back now mind you these routing information that this these routes will never actually be used for routing but you know what they might populate the databases here and this is a looped information not to mention that the similar thing may have happened in the other direction r2 here or border - may have injected this information back so now we are going to get even more information in the database here and this information makes its way back so we somehow need to find a way to prevent two loops happening here because what we have now is really not a simple hub-and-spoke but what we have now is a redundant hub-and-spoke environment now when you have redundant hub-and-spoke environment unless it is are all the same interfaces how do you prevent two loops well you prevent two loops by taking the routes and preventing them from being advertised out of the same interface they are received on so what we need to do here is implement a split horizon so when people ask you does OSPF implement split horizon of course you're going to answer no because at the interface level OSPF does not implement split horizon but at the area level in order to prevent this problem from happening we may need to do something that looks like a split horizon so what can we actually do to address and fix this problem so let's take a look at that I'm going to paste this in and again please bear with me until I clean it up just a little bit maybe I should really have more efficient copy paste buffer than this I'm efficient when it comes to configuring stuff but when it comes to drawing stuff I'm really not oops really not that good at it so what we need to do here is we need to prevent this information that is injected by this border one from ever leaving the area so this is where this information needs to stop also information injected by this router here needs to be prevented from leaving this area here and OSPF does that so on the border routers what we are going to have is a split horizon rule Type three received from area zero will be injected will be summarized by that I mean new type three will be generated into non zero area information or actually type 3 received from nonzero area will not be used for routing that basically means that information that was injected here by r2 will be in database here will be in database here database here it needs to be in the database on this Roderick cannot reject it because the database must be in sync in the whole area but this router will not use this information for routing if it doesn't use it for routing it cannot calculate the path for it why because if it's not in a routing table there will be no calculated cost for it if there is no calculated cost for it that means that this route will never leave this area same thing happens with this information here so this is the split horizon rule that OSPF implements and prevents the loops in a multi area environment this is very very important because there will be situations that some router appears to be an ABR but it actually is not an ABR so let me give you that example so let's say that we have a router r1 that sits between two areas the number of interfaces or membership in these areas is relevant so let's say that this is area 1 and this is area 2 if you have this situation this router in OSPF in sorry in I in iOS is not an ABR now this router is not an ABR because it's connected to area 0 in area 2 which are non backbone areas there will be no summarization of information between these two areas so this router will have a database in area 2 and will have database in area 1 but there will be no information exchanged between these two areas because in iOS the routers cannot summarize between nan backbone areas they can only summarize the LSA information between non backbone and backbone area and between backbone area and the non backbone area so this direct summarization in iOS is not an ABR now I'm saying not in iOS because for example in Juno's it is so in Juno's this rudder will be for example considered to be in APR but there are still carriers with this and I'm just putting an asterisk there just to annoy it this is an iOS specific behavior there is actually an RFC that is written what makes a router and a B are in different implementations of iOS because RFC is rather vague on the subject of what makes an AVR but in iOS this is not an ABR which brings me to the next point the next point here is let's say that we had a network that looks and I believe I already have that thing so let me bring one of the older slides just have to find we've done so many in between so that the one that I need might be a little bit lost I will need to clean it up so maybe I'll be just faster actually rejoined yeah we'll definitely be faster enjoyed so R 1 R 2 R 3 and then R 4 and let's sit here that this is area one this is area two and this is area three and let's say that we have a network n that exists here we know that this will be advertised as type one LSA and it will exist in database here but because this is area 1 and this is area 2 there will be no summarization of this information going through so no routers in area 2 can actually reach any networks from area 1 so this is a known thing but what if for example this was area 0 and this was area 0 what is the situation there well let's take a look we know that this will information be sent as type 1 LS 8 will exist in this database here this router now is an ABR right so this router here is an ABR which means that this information will be actually summarized and sent to RT which will now have this information in its database here but take a look this is a known backbone area and this was a type 3 LSA received in a non backbone area what is r3 going to do with it it cannot use this information for anything which means that these guys here in area 0 cannot reach the node cannot reach the destinations in another area 0 so this here requires effects now the fix of course stems from another requirement in OSPF and the requirement that we are failing to meet here is that area 0 must be continuous area that means that we must have an uninterrupted connections between area zero we cannot have split brain like this so this is supposed to be the exact same database here but it is not because this router here cannot inject the type 3 LSA from an on backbone area back into the backbone area because of the split horizon room so how do we fix this well the fix for this well there are multiple fixes somehow we need to extend area zero across area two and one of the possible solutions is to use something called the virtual link so what we might need to do here is I can just use this what we might need to do here is the virtual link now virtual link is IP unnumbered point-to-point link which is always in area 0 so you cannot have a virtual link that is in a non zero area now this might sound confusing because the actual syntax for configuring virtual link is router OSPF let's say one and then you say area in our case this was area two virtual link and then you specify the router ID of the destination so what am I saying that virtual link is always in area zero when this one is in area two well area two in this case is something that we call the transit area so we are telling our router create the virtual link using area two as the transit area but the actual virtual and so let's say that we had r1 and then we had any number of routers here and we had r2 so let's say that this here was area zero this was area zero this is area two what we are basically telling our routers in area to create this virtual connection between r1 and r2 but this virtual interface that gets created here and created here will actually be in area zero creating one contiguous area zero so that the database on this side here and the database on this side here can actually be in sync so this is what the virtual links do now another solution that you can use and I'm not going to spend too much time on it is that you can actually use GRE tunnel for the same purpose but you just be up a little bit more but don't think about it so you can use GRE tunnel about it as well and with GRE tunnel you are not restricted to actually being it unnumbered you can put IP addresses on the interfaces but if you decide to use the virtual and this is what happens also sometimes especially in the CGI lab environment what you might encounter is the situation so here I'm going to use a little bit more routers than just for is that you have for example now to be an extreme case that you have area 0 here in area 0 here whether this is area 1 2 & 3 now how do you interconnect these two areas Eros well if you want to use virtual links what you actually need to do is you need to set up a cascade of virtual links so you need the virtual in here you will need a virtual in here you will need a virtual in here and then you would have native connectivity here and here inside area 0 but basically what you will have is one contiguous area 0 in this case now keep in mind that in any of these areas so let me up use this area as an example you don't have to have a direct connection between these routers you can have any of routers in between but you are always going to be building virtual link between ABR so this is an ABR and this is an ABR you cannot cross multiple areas with virtual links and the reason for that is that the destination of the virtual link is not is not an IP address but it's a router ID so this is very very very significant piece of information one of the questions that I received is when creating a virtual link between loopback interfaces of two routers does one of them need to be in area 0 I'm going to repeat this again the virtual links are built between router IDs they are not built between the interfaces so it doesn't matter in which area interface that borrowed the number 2 router ID resides because the virtual link is not built between that interface and another interface it's built between router IDs so rato ID is not an IP address so if you want to build a virtual in-between router r1 and router r2 and r1 has looked like interface in area 19 and that loop back address borrowed its IP address for router ID that's irrelevant because router ID exists in all areas because router ID is unique per process not per area so the answer is now you don't need to have loopback in area 0 it's because it's relevant another question was if I have this cascading set up that that I need to do if I have these multiple areas and if I wanted to build a GRE tunnel between let's say this ABR here and this ABR here that would require no less work than using cascading virtual links well that depends because you see for GRE tunnel to come up between this point and this point here for GRE to come up all you need is to have the route to an end point so you can have static routes you can have ER GRP routes you can have any means as long as this router somehow can ping this router here not necessarily true or SPF but if you have some other ways of reaching it you could actually build a GRE tunnel by passing these two here or by passing area to complexity here so with GRE tunnel you are not necessarily bound to this cascading setup but you would need to have some external ways of knowing the route but for virtual links you cannot use these shortcuts for virtual links you actually have to do it between the ABR's because again virtual link endpoints are not I P addresses they are router IDs but for gr is their IP addresses and if IP address is reachable you don't care how you can build the GRE tunnel and then if you have the GRE tunnel you can run OSPF over it and you can actually interconnect these two areas another question is if in a virtual link scenario so unrelated to GRE if I need to build the the the virtual links between the router IDs do router IDs need to be routable again they are not IP addresses they are not routable they are not pinging able not necessarily but if you have valid adjacencies in one area you can assume that from OSPF perspective the OSPF will know how to reach that router ID that said this was actually an excellent question why is that because even if you have router IDs so let's say that we have router here and we have router here and we have something in between them right and what I want to do is I want to build a virtual link between these two routers so these are the ABR's so this is area 0 here this is area 1 and this is area 0 here now when I'm building this virtual link what I'm configuring is router OSPF let's say that this is a br - and this is a br one so I'm going to say router OSPF one and I'm going to say area one virtual link and let's say ABR - now what I'm entering here is the router ID but you know what the actual IP packet needs to send be sent here so what router ABR one is going to do when it sends the hello so this is the hello that needs to be censored this is an actual IP packet an IP packet must have the source IP address and must have the destination IP address so what ABR one needs to determine is which source IP address to use and which destination IP address to use so the router ID itself is not an IP address it's not routable but we need to find at least one routable IP address inside this area on this router in order to be able to reach it which can open a whole new set of problems if this here was for example IP unnumbered link now if this was an IP unnumbered link that it borrowed the IP address from another area your virtual in here will never come up in that case what you need to do is you will probably need to build a jury tunnel up to here or maybe virtual link up to here and then you would probably need to build a GRE tunnel here and hoping that this year was a PPP link because if it was hdl-c it wouldn't work but because if it was PPP you could build a GRE tunnel because in IP CPU will have that weird / 32 for your neighbor now there is a blog post there is an article that I have written on this subject you can find it on blog that IP expert calm its OSPF and IP n number links it actually describes this scenario and why it's so complicated and complex at this point you may not remember that we have actually been working on some OSPF table but let me remind you about that so this is the table of our LSA types now I went on a tangent explaining what type three is how the OSPF works in in in their area environment so knowing all that stuff type three lsas are actually going to be generated by ABR routers or the area border routers and the data content will be summarized information contained in type one and optionally type two LS s so this is the information we summarized and let's see if I can fit this in one like so now the flooding scope is going to be an area

Channel: CCIEORDIE.COM

Views: 1,403

Keywords:

Id: dm7p1V9ihh0

Length: 58min 22sec (3502 seconds)

Published: Wed Jan 31 2018