WEBVTT

0:00:03.040000 --> 0:00:07.140000
 Hello and welcome to this video, which
 is a refresher for the CCNA 200

0:00:07.140000 --> 0:00:11.920000
-301 on Open Shortest
 Path First, OSPF.

0:00:11.920000 --> 0:00:18.100000
 Let's talk about OSPF is an interior
 gateway routing protocol.

0:00:18.100000 --> 0:00:21.820000
 So it was designed to operate
 within an autonomous system.

0:00:21.820000 --> 0:00:24.840000
 Just think of that as being synonymous
 with within a company's network

0:00:24.840000 --> 0:00:27.460000
 to exchange routes within
 that autonomous system.

0:00:27.460000 --> 0:00:31.760000
 It is an open standard, so
 it was specified in RFC.

0:00:31.760000 --> 0:00:35.920000
 It's always been open, so any vendor's
 equipment should be able to support

0:00:35.920000 --> 0:00:41.200000
 OSPF. It uses the shortest path first
 algorithm, otherwise known as the

0:00:41.200000 --> 0:00:46.040000
 Dijkstra algorithm to compute the
 shortest path to any destination.

0:00:46.040000 --> 0:00:50.240000
 One thing that makes OSPF interesting
 when it comes to routing protocols

0:00:50.240000 --> 0:00:54.680000
 is that routing protocols generally
 fall into two different categories.

0:00:54.680000 --> 0:00:59.920000
 There are those categories where a router
 will learn of routes from its

0:00:59.920000 --> 0:01:03.660000
 directly connected neighbor, so from another
 router that's directly physically

0:01:03.660000 --> 0:01:06.000000
 connected to it across a cable.

0:01:06.000000 --> 0:01:09.680000
 And so a learner of routes, and it'll
 use that neighbor as the next hop.

0:01:09.680000 --> 0:01:14.460000
 But it won't really know what the topology
 looks like behind that neighbor.

0:01:14.460000 --> 0:01:17.860000
 It'd be sort of like me telling you,
 hey, you don't know what's on the

0:01:17.860000 --> 0:01:19.900000
 other side of this curtain, what's on
 the other side of this wall right

0:01:19.900000 --> 0:01:25.860000
 here. But if you go through me, there's
 like 50 different rooms back there.

0:01:25.860000 --> 0:01:28.680000
 Well, you don't know what the size
 and shape of those rooms are.

0:01:28.680000 --> 0:01:29.880000
 You don't know what the layout is.


0:01:29.880000 --> 0:01:34.580000
 You just know that if you ever need
 to send a message to any of those

0:01:34.580000 --> 0:01:38.640000
 rooms, just give that message to
 me, and I'll send it back there.

0:01:38.640000 --> 0:01:41.120000
 So what's behind that curtain is
 pretty much invisible to you.

0:01:41.120000 --> 0:01:44.600000
 You just know there's 50 different rooms
 back there, maybe the names of

0:01:44.600000 --> 0:01:46.120000
 each one of those rooms.

0:01:46.120000 --> 0:01:47.660000
 That'd be synonymous to a route.

0:01:47.660000 --> 0:01:50.360000
 Well, OSPF takes a
 different approach.

0:01:50.360000 --> 0:01:55.500000
 OSPF says, not only do you know about
 me and this connected network that

0:01:55.500000 --> 0:01:59.240000
 we have, you know exactly what everything
 looks like behind me.

0:01:59.240000 --> 0:02:03.940000
 You've got a full topology of what this
 building looks like, how the rooms

0:02:03.940000 --> 0:02:08.120000
 are interconnected, their sizes, their
 shapes, where their doors are.

0:02:08.120000 --> 0:02:09.700000
 That's like what OSPF is.

0:02:09.700000 --> 0:02:14.520000
 OSPF router learns of what the whole
 topology looks like, and then decides

0:02:14.520000 --> 0:02:21.500000
 for itself what the best neighbor is
 to use to get to a particular route.

0:02:21.500000 --> 0:02:24.380000
 So OSPF forms neighbor
 relationships.

0:02:24.380000 --> 0:02:30.120000
 Before any information, before any prefix
 or topology information is exchanged

0:02:30.120000 --> 0:02:34.980000
 between a pair of routers, they have to
 form what's called an OSPF adjacency,

0:02:34.980000 --> 0:02:37.860000
 what we sometimes call an OSPF
 neighbor relationship.

0:02:37.860000 --> 0:02:41.800000
 And they do that by exchanging special
 type of OSPF messages called hello

0:02:41.800000 --> 0:02:47.040000
 packets. As you can see here on the
 vast majority of network segments,

0:02:47.040000 --> 0:02:51.380000
 hello packets go out every 10 seconds
 and they have a hold timer of 40

0:02:51.380000 --> 0:02:55.720000
 seconds. And yes, you will want to memorize
 those timers, especially if

0:02:55.720000 --> 0:02:57.380000
 you're taking any exam.

0:02:57.380000 --> 0:03:03.020000
 Now, it's more than just a simple one
 to exchange of hellos for routers

0:03:03.020000 --> 0:03:06.040000
 to say, oh, okay, let me share
 my information with you.

0:03:06.040000 --> 0:03:10.280000
 In some protocols like EIGRP,
 it is about that simple.

0:03:10.280000 --> 0:03:14.440000
 But in OSPF, there's about five or
 six different steps a router has to

0:03:14.440000 --> 0:03:18.380000
 go through before it can become fully
 synchronized with another neighbor

0:03:18.380000 --> 0:03:20.720000
 and have all the information.

0:03:20.720000 --> 0:03:26.720000
 OSPF works based on an area hierarchy
 which minimizes LSA flooding.

0:03:26.720000 --> 0:03:28.000000
 What is that talking about?

0:03:28.000000 --> 0:03:35.180000
 All right. So in the world of OSPF,
 when I generate information that I'm

0:03:35.180000 --> 0:03:40.000000
 going to give to you, let's say you are
 my OSPF neighbor, my primary objective

0:03:40.000000 --> 0:03:45.280000
 in generating this packet called a
 link state advertisement in LSA is

0:03:45.280000 --> 0:03:47.600000
 not to give you routing
 information.

0:03:47.600000 --> 0:03:49.400000
 That's not the primary
 purpose of it.

0:03:49.400000 --> 0:03:53.240000
 The primary purpose of it is to give
 you descriptive information about

0:03:53.240000 --> 0:03:58.560000
 me, you know, my name, what links I have,
 what the cost is of those links,

0:03:58.560000 --> 0:03:59.700000
 what type of links they are.

0:03:59.700000 --> 0:04:00.720000
 Are they point to point?

0:04:00.720000 --> 0:04:02.720000
 Are they something else?

0:04:02.720000 --> 0:04:06.840000
 Because the goal is I want to give
 you enough information that you can

0:04:06.840000 --> 0:04:11.960000
 plug me into your map that you know
 about me, you know who I'm connected

0:04:11.960000 --> 0:04:14.200000
 to because I'm telling you that.

0:04:14.200000 --> 0:04:18.600000
 And so all these routers will create
 link state advertisements and flood

0:04:18.600000 --> 0:04:22.600000
 them. So the link state advertisement
 I give to you, you will put it in

0:04:22.600000 --> 0:04:26.340000
 your database, you'll store it, but you'll
 also copy it and keep it going

0:04:26.340000 --> 0:04:29.680000
 to the neighbor behind you and
 the neighbor behind him.

0:04:29.680000 --> 0:04:35.880000
 So within OSPF area, everybody is
 going to learn about everything.

0:04:35.880000 --> 0:04:37.900000
 Now, what's this concept of areas?


0:04:37.900000 --> 0:04:39.960000
 Let's just do a quick
 review of that.

0:04:39.960000 --> 0:04:46.520000
 So one of the rules of OSPF is that
 there are several different kinds

0:04:46.520000 --> 0:04:52.360000
 of LSA's. And at the CCNA level, you
 don't have to know the names and

0:04:52.360000 --> 0:04:55.920000
 numbers and functionality of
 every single kind of LSA.

0:04:55.920000 --> 0:04:59.440000
 But the fundamental one you should
 know about, which I've mentioned in

0:04:59.440000 --> 0:05:05.880000
 other videos, is called
 the type one router LSA.

0:05:05.880000 --> 0:05:08.680000
 This is like what I
 was just describing.

0:05:08.680000 --> 0:05:12.460000
 The main purpose of the router LSA is
 so that all the routers can learn

0:05:12.460000 --> 0:05:17.580000
 about me, they can learn my links, they
 can learn who I'm connected to,

0:05:17.580000 --> 0:05:21.300000
 and then by using that, and the LSA's
 they get from other routers, they

0:05:21.300000 --> 0:05:23.560000
 can piece together
 a jigsaw puzzle.

0:05:23.560000 --> 0:05:27.160000
 Well, one of the rules of OSPF says,
 let's say that your jigsaw puzzle

0:05:27.160000 --> 0:05:31.080000
 is complete. You know the full topology,
 you've figured out the full map

0:05:31.080000 --> 0:05:34.260000
 or what we call the tree of where everybody
 is and how they're connected

0:05:34.260000 --> 0:05:37.420000
 to in your CPU in your mind.

0:05:37.420000 --> 0:05:40.940000
 And then all of a sudden,
 you get a new router LSA.

0:05:40.940000 --> 0:05:45.060000
 That topology has been nice and stable
 for the last day or week or month,

0:05:45.060000 --> 0:05:47.760000
 and all of a sudden, now
 a new LSA comes in.

0:05:47.760000 --> 0:05:51.460000
 Maybe about a brand new router, maybe
 about an existing router whose link

0:05:51.460000 --> 0:05:57.080000
 has gone down. Well, the rules of OSPF
 say, if I get a new type one router

0:05:57.080000 --> 0:06:02.280000
 LSA, I have to break apart my tree,
 which took a lot of memory and CPU

0:06:02.280000 --> 0:06:04.960000
 resources to build
 in the first place.

0:06:04.960000 --> 0:06:07.500000
 Now think about a jigsaw puzzle, it's
 not easy to build those things,

0:06:07.500000 --> 0:06:12.740000
 right? You got to tear that thing apart
 and then rebuild it again, factoring

0:06:12.740000 --> 0:06:14.980000
 in that new LSA that you just got.


0:06:14.980000 --> 0:06:20.200000
 So the result of this is, if there's
 any instability in your network,

0:06:20.200000 --> 0:06:23.460000
 if there's any link or link somewhere
 that are flapping going down, going

0:06:23.460000 --> 0:06:29.240000
 up, going down, going up, that's going
 to be intrusive on all the OSPF

0:06:29.240000 --> 0:06:32.800000
 routers, because every time it
 goes down, a new LSA goes out.

0:06:32.800000 --> 0:06:35.320000
 So the router that had that link that
 just went down, he has to create

0:06:35.320000 --> 0:06:38.200000
 a new router LSA saying, hey,
 everybody remember me?

0:06:38.200000 --> 0:06:42.520000
 This link I told you about, it's gone,
 tear apart your tree, rebuild it,

0:06:42.520000 --> 0:06:43.860000
 factoring that information.

0:06:43.860000 --> 0:06:45.780000
 Link comes up. Hey, everybody
 remember about me?

0:06:45.780000 --> 0:06:49.880000
 That link just came up and you can
 see that would be very intrusive.

0:06:49.880000 --> 0:06:53.980000
 So the developers of OSPF who created
 this, they said, we're going to

0:06:53.980000 --> 0:06:56.040000
 give people some options
 right here.

0:06:56.040000 --> 0:07:01.060000
 So what we're going to do is we're going
 to say that when you configure

0:07:01.060000 --> 0:07:07.320000
 OSPF that a link, which is basically
 a cable and network interface, is

0:07:07.320000 --> 0:07:10.160000
 considered to be in an area.

0:07:10.160000 --> 0:07:15.740000
 Now all the routers, go ahead and draw
 this right here, you could, if

0:07:15.740000 --> 0:07:20.480000
 you wanted to, design your network
 in such a way that all the routers

0:07:20.480000 --> 0:07:27.120000
 and all their links like this right
 here, we're all in one big area.

0:07:27.120000 --> 0:07:28.720000
 You could do that.

0:07:28.720000 --> 0:07:30.600000
 That's perfectly fine.

0:07:30.600000 --> 0:07:33.760000
 But OSPF does require that
 you select an area number.

0:07:33.760000 --> 0:07:37.460000
 So if you just have one area,
 that can be any area you want.

0:07:37.460000 --> 0:07:40.200000
 I could make that
 one area area 17.

0:07:40.200000 --> 0:07:45.100000
 Now it's probably a good design
 practice to make it area zero.

0:07:45.100000 --> 0:07:47.400000
 And I'll talk about why that
 isn't just a second.

0:07:47.400000 --> 0:07:50.980000
 But the takeaway from here is you could
 put all your links in one area

0:07:50.980000 --> 0:07:54.240000
 if you wanted to, and that area
 number could be the same.

0:07:54.240000 --> 0:07:57.360000
 So one thing that's very critical is
 that when two routers are exchanging

0:07:57.360000 --> 0:08:01.920000
 LSA's, they're going to tell each other
 what area they believe their link

0:08:01.920000 --> 0:08:07.100000
 is in. If they don't agree on that,
 they will not become OSPF neighbors.

0:08:07.100000 --> 0:08:09.900000
 They have to agree that that
 link is in a common area.

0:08:09.900000 --> 0:08:12.560000
 Now here's the downside.

0:08:12.560000 --> 0:08:15.880000
 If I put all my routers in one area
 like this, and hey, if I've only got

0:08:15.880000 --> 0:08:18.720000
 five or six routers like
 this, go for it.

0:08:18.720000 --> 0:08:20.040000
 It's not going to hurt anything.

0:08:20.040000 --> 0:08:24.760000
 But imagine if I had 100 routers, well,
 now we've got a greater likelihood

0:08:24.760000 --> 0:08:28.720000
 of instability. The more routers, the
 more links you have, the greater

0:08:28.720000 --> 0:08:32.460000
 likelihood that somewhere, some link
 is going to flap and go down, which

0:08:32.460000 --> 0:08:34.940000
 is going to cause a problem
 for everybody.

0:08:34.940000 --> 0:08:39.660000
 So the developers of OSPF said, hey,
 why don't you do this instead?

0:08:39.660000 --> 0:08:44.280000
 If you want to, and if you think there
 might be some instability, instead

0:08:44.280000 --> 0:08:50.060000
 of putting all your routers in one big
 area, why don't you subdivide it

0:08:50.060000 --> 0:08:52.580000
 into two or more areas?

0:08:52.580000 --> 0:08:53.900000
 So why don't you do
 something like this?

0:08:53.900000 --> 0:08:58.960000
 Why don't we say that these links are
 going to be in one area, and we

0:08:58.960000 --> 0:09:04.640000
 can say that these other links right
 here are in a different area.

0:09:04.640000 --> 0:09:08.940000
 So now, if there's a difference between
 the two any instability in this

0:09:08.940000 --> 0:09:13.020000
 area here, this blue area, it's only
 going to affect those four routers.

0:09:13.020000 --> 0:09:16.680000
 It will not affect the two routers
 on the left because those type one

0:09:16.680000 --> 0:09:20.680000
 router LSAs I talked to you about that's
 used to build the topology, those

0:09:20.680000 --> 0:09:25.020000
 are only allowed to be
 flooded within an area.

0:09:25.020000 --> 0:09:26.660000
 They can go no further.

0:09:26.660000 --> 0:09:28.580000
 So they're going to be flooded
 around through here.

0:09:28.580000 --> 0:09:34.720000
 But once they get to this guy who's
 bordering two areas, they stop at

0:09:34.720000 --> 0:09:38.920000
 him. So if there's any instability,
 it will only affect the CPUs of the

0:09:38.920000 --> 0:09:40.740000
 routers in that blue area.

0:09:40.740000 --> 0:09:43.100000
 It will not affect the routers
 in the purple area.

0:09:43.100000 --> 0:09:46.000000
 Now, you might be wondering, well, wait
 a second, if those LSAs only get

0:09:46.000000 --> 0:09:49.320000
 to that guy and they stop, how is the
 purple area ever going to learn

0:09:49.320000 --> 0:09:52.660000
 about the routes in the blue area?


0:09:52.660000 --> 0:09:56.420000
 And vice versa. Well, there's a different
 kind of an LSA that is doing

0:09:56.420000 --> 0:10:00.140000
 that. There's a special kind of an LSA
 that this guy right here creates

0:10:00.140000 --> 0:10:04.500000
 to advertise routes from one
 area into another area.

0:10:04.500000 --> 0:10:08.280000
 But that LSA does not
 affect the tree.

0:10:08.280000 --> 0:10:11.580000
 Those types of LSAs can come
 and go all day long.

0:10:11.580000 --> 0:10:14.440000
 And if your tree is nice and stable,
 it won't affect those.

0:10:14.440000 --> 0:10:17.400000
 It does not. So when you receive those
 LSAs, you don't have to break apart

0:10:17.400000 --> 0:10:19.560000
 your tree to account for those.

0:10:19.560000 --> 0:10:23.300000
 Only for these type one router LSAs do
 you have to take that drastic action.

0:10:23.300000 --> 0:10:29.220000
 Now, when you do have two or more areas,
 now we do have a restriction.

0:10:29.220000 --> 0:10:32.820000
 One of those areas has to be the backbone
 area, which is what we call

0:10:32.820000 --> 0:10:38.380000
 area zero. The other area, it
 can be whatever you want.

0:10:38.380000 --> 0:10:43.900000
 Area one. So usually what I tell people
 is, if you're going to design

0:10:43.900000 --> 0:10:48.680000
 an OSPA network that has two or more areas,
 you should think of your design

0:10:48.680000 --> 0:10:53.660000
 as sort of representing like
 a flower, like this.

0:10:53.660000 --> 0:11:00.620000
 Okay, where this is area zero, and
 these are all different areas, you

0:11:00.620000 --> 0:11:04.080000
 can make them whatever
 numbers you want.

0:11:04.080000 --> 0:11:10.900000
 And you have these routers right here
 that are connecting those areas.

0:11:10.900000 --> 0:11:14.960000
 You could have one router
 connecting both like that.

0:11:14.960000 --> 0:11:21.260000
 And these routers, a router that has
 a link in area zero and a link in

0:11:21.260000 --> 0:11:24.880000
 something else, we call those
 area border routers.

0:11:24.880000 --> 0:11:29.120000
 They're performing that special job
 of creating these special kinds of

0:11:29.120000 --> 0:11:32.500000
 LSAs that advertise routes from
 one area into the other.

0:11:32.500000 --> 0:11:34.920000
 Only area border routers
 can do that.

0:11:34.920000 --> 0:11:39.360000
 Now, you see, if I forgot about that
 rule, and I had another area out

0:11:39.360000 --> 0:11:44.460000
 here, like area 55, and here was a router
 right there with a link in area

0:11:44.460000 --> 0:11:48.640000
 55 and a link in area 18, that would
 not be an area border router.

0:11:48.640000 --> 0:11:51.540000
 He would know that he's not an area
 border router because he would say,

0:11:51.540000 --> 0:11:53.620000
 I don't have any links
 in area zero.

0:11:53.620000 --> 0:11:57.540000
 So he would learn about the routes in
 area 55, but he would not be allowed

0:11:57.540000 --> 0:12:01.000000
 to forward those routes
 into area 18.

0:12:01.000000 --> 0:12:03.520000
 He couldn't do that because
 it's not his job.

0:12:03.520000 --> 0:12:08.520000
 So that's why it's a better idea when
 designing an OSPF network to design

0:12:08.520000 --> 0:12:12.620000
 it in like this flower type of shape
 here, where each pedal is connecting

0:12:12.620000 --> 0:12:14.860000
 into area zero in the center.

0:12:14.860000 --> 0:12:22.760000
 So as this slide mentions here,
 minimizes LSA flooding.

0:12:22.760000 --> 0:12:26.240000
 And this is one of the big benefits of
 dividing your network into multiple

0:12:26.240000 --> 0:12:30.940000
 areas, because if there's any instability
 in one area, the LSAs that are

0:12:30.940000 --> 0:12:36.120000
 flooded that carry that instability
 are not propagated into other areas.

0:12:36.120000 --> 0:12:45.080000
 So the metric that OSPF uses, so when
 an OSPF router learns via LSAs,

0:12:45.080000 --> 0:12:50.220000
 two or more paths to the exact same
 network, it has to use a metric to

0:12:50.220000 --> 0:12:51.240000
 figure out the best route.

0:12:51.240000 --> 0:12:55.140000
 And we know that in all routing protocols,
 the lower the metric, the better

0:12:55.140000 --> 0:12:59.220000
 the route. Well, the metric
 that OSPF uses is cost.

0:12:59.220000 --> 0:13:00.060000
 That's the name of it.

0:13:00.060000 --> 0:13:04.440000
 And like spanning tree, OSPF cost
 is based on interface bandwidth.

0:13:04.440000 --> 0:13:09.100000
 The greater the bandwidth or the faster
 the interface, the lower the cost.

0:13:09.100000 --> 0:13:15.760000
 So the best path to any given network
 is the path with the lowest cumulative

0:13:15.760000 --> 0:13:19.840000
 cost, just like spanning tree decides
 its best path to get to the root

0:13:19.840000 --> 0:13:22.580000
 bridge. Now the formulas
 are different.

0:13:22.580000 --> 0:13:26.540000
 Actually, the formula for OSPF is
 actually very easy to remember.

0:13:26.540000 --> 0:13:35.660000
 The formula for OSPF is you just take
 100 million and divide it by the

0:13:35.660000 --> 0:13:42.160000
 interface bandwidth
 in bits per second.

0:13:42.160000 --> 0:13:49.340000
 So for example, if I had a fast ethernet
 interface, well, fast ethernet

0:13:49.340000 --> 0:13:52.100000
 is 100 million bits per second.

0:13:52.100000 --> 0:13:53.260000
 That's fast ethernet.

0:13:53.260000 --> 0:13:58.200000
 So we would end up having 100 million
 divided by our bandwidth, which

0:13:58.200000 --> 0:14:02.940000
 is also 100 million, which would
 give us a cost of one.

0:14:02.940000 --> 0:14:05.560000
 That's the cost of a fast
 ethernet interface.

0:14:05.560000 --> 0:14:14.780000
 Slower interfaces would
 have bigger costs.

0:14:14.780000 --> 0:14:18.860000
 And costs can be configured explicitly
 or by affecting or by changing

0:14:18.860000 --> 0:14:20.800000
 the interface bandwidth.

0:14:20.800000 --> 0:14:27.140000
 So you can go, I'll go ahead and show
 you right here, go on to a router.

0:14:27.140000 --> 0:14:33.260000
 Now this guy is not running OSPF yet,
 but if I go on to an interface,

0:14:33.260000 --> 0:14:38.540000
 for example, if I do show interface
 gigabit zero slash one.

0:14:38.540000 --> 0:14:43.840000
 Okay, see how this interface
 has a bandwidth?

0:14:43.840000 --> 0:14:46.480000
 Now this is in kilobits
 per second.

0:14:46.480000 --> 0:14:51.120000
 So we'd have to add three zeros to
 this to get the actual bandwidth in

0:14:51.120000 --> 0:14:52.460000
 bits per second.

0:14:52.460000 --> 0:14:59.580000
 But this number with three extra zeros,
 as the, I always forget as the

0:14:59.580000 --> 0:15:02.780000
 denominator, the one on the bottom,
 or the numerator, the one, anyway,

0:15:02.780000 --> 0:15:05.980000
 in a division, the one on the bottom,
 that's where the bandwidth would

0:15:05.980000 --> 0:15:10.700000
 go. So if we have 100 million divided
 by this, that's what your OSPF cost

0:15:10.700000 --> 0:15:14.420000
 would be. As a matter of fact, let's
 just go ahead and do that.

0:15:14.420000 --> 0:15:16.960000
 Interface gig zero slash one.

0:15:16.960000 --> 0:15:19.620000
 Do we have an IP address on there?


0:15:19.620000 --> 0:15:21.620000
 I don't think so.

0:15:21.620000 --> 0:15:29.240000
 So let's just put one IP address.

0:15:29.240000 --> 0:15:32.380000
 All right, IP OSPF one area zero.

0:15:32.380000 --> 0:15:34.640000
 We'll get to the commands
 for this in just a second.

0:15:34.640000 --> 0:15:37.740000
 I don't want you to focus
 on the command just yet.

0:15:37.740000 --> 0:15:39.560000
 But this is what I
 want you to see.

0:15:39.560000 --> 0:15:50.400000
 So when I do the commands show IP OSPF
 interface for that interface, it

0:15:50.400000 --> 0:15:54.100000
 shows me right here
 the cost is one.

0:15:54.100000 --> 0:15:57.100000
 Now you said, but wait a second, Keith,
 you said a fast ethernet was one.

0:15:57.100000 --> 0:15:58.040000
 This is gigabit.

0:15:58.040000 --> 0:15:59.740000
 Well, here's the problem.

0:15:59.740000 --> 0:16:02.220000
 Fast ethernet is one, right?

0:16:02.220000 --> 0:16:04.200000
 Fast ethernet is 100 million
 bits per second.

0:16:04.200000 --> 0:16:09.780000
 And actually, everything from fast ethernet
 on up, gigabit, 10 gigabit,

0:16:09.780000 --> 0:16:13.520000
 40 gigabit, they're all going to come
 up with a cost of one as far as

0:16:13.520000 --> 0:16:15.240000
 OSPF is concerned.

0:16:15.240000 --> 0:16:19.140000
 Now there is a command you can use
 to change that, but just know that

0:16:19.140000 --> 0:16:20.900000
 that's the default behavior.

0:16:20.900000 --> 0:16:22.360000
 So the cost is one.

0:16:22.360000 --> 0:16:24.560000
 Now watch how I can change this.

0:16:24.560000 --> 0:16:27.680000
 I can go back onto that interface.


0:16:27.680000 --> 0:16:30.340000
 One thing I could do is I could
 change the bandwidth.

0:16:30.340000 --> 0:16:34.460000
 So let's give it some
 really low bandwidth.

0:16:34.460000 --> 0:16:38.860000
 How about 128 kilobits per second?


0:16:38.860000 --> 0:16:41.580000
 It's almost like a dial
 up line at this point.

0:16:41.580000 --> 0:16:48.520000
 Now if I do show IP OSPF interface
 gigabit zero slash one, you need to

0:16:48.520000 --> 0:16:50.280000
 proceed that with a do.

0:16:50.280000 --> 0:16:58.280000
 Look at that. Now it increased the cost
 to 781 because I made the bandwidth

0:16:58.280000 --> 0:17:01.600000
 a lot slower. It really
 bumped up the cost.

0:17:01.600000 --> 0:17:07.900000
 So this interface is not as good from
 the perspective of OSPF anymore.

0:17:07.900000 --> 0:17:14.800000
 Or I could actually use this command,
 IP OSPF cost and then explicitly

0:17:14.800000 --> 0:17:17.940000
 give it a cost. How about 27?

0:17:17.940000 --> 0:17:22.200000
 So now he's no longer looking
 at the bandwidth at all.

0:17:22.200000 --> 0:17:30.380000
 Now he's looking directly
 at what I did.

0:17:30.380000 --> 0:17:33.560000
 So that's what that is talking about
 there at the bottom that you can

0:17:33.560000 --> 0:17:38.200000
 configure explicitly or or modify the
 bandwidth command to have that cost

0:17:38.200000 --> 0:17:42.620000
 changed. All right.

0:17:42.620000 --> 0:17:46.040000
 There are several different OSPF packet
 types that OSPF uses to accomplish

0:17:46.040000 --> 0:17:49.460000
 its goals. We talked about
 the OSPF hello packet.

0:17:49.460000 --> 0:17:54.600000
 So this is actually also showing the
 packets are exchanged when we're

0:17:54.600000 --> 0:17:56.020000
 building a neighbor relationship.

0:17:56.020000 --> 0:17:58.440000
 So let's just walk through
 this process.

0:17:58.440000 --> 0:18:02.480000
 So first thing that happens is you and
 I, let's just say you're a router

0:18:02.480000 --> 0:18:05.760000
 and I'm a router and we're directly
 connected across a link.

0:18:05.760000 --> 0:18:09.120000
 First thing is we have to exchange hello
 packets and there's some parameters

0:18:09.120000 --> 0:18:11.720000
 in those hello packets
 that have to match.

0:18:11.720000 --> 0:18:14.260000
 And we'll talk about that
 in just a moment.

0:18:14.260000 --> 0:18:18.400000
 Assuming those parameters match, then
 we exchange these things called

0:18:18.400000 --> 0:18:20.240000
 database descriptor packets.

0:18:20.240000 --> 0:18:22.360000
 What is that and why do we do it?

0:18:22.360000 --> 0:18:28.400000
 Well, our objective here, our objective
 is that if you and I are in the

0:18:28.400000 --> 0:18:32.320000
 same area, which we need to be otherwise
 we're not going to be neighbors,

0:18:32.320000 --> 0:18:37.680000
 ultimately, whatever LSAs you have
 in your database have to match mine

0:18:37.680000 --> 0:18:41.280000
 and vice versa. We call
 that synchronization.

0:18:41.280000 --> 0:18:45.300000
 You and I need to be synchronized as far
 as our LSA databases are concerned.

0:18:45.300000 --> 0:18:48.500000
 Well, at the beginning, we don't know
 what each other has and we don't

0:18:48.500000 --> 0:18:50.560000
 know for missing some stuff.

0:18:50.560000 --> 0:18:54.420000
 So the purpose of the database descriptor
 exchange and that's a type of

0:18:54.420000 --> 0:18:59.440000
 OSPF packet is for me to give you like
 a high level overview of the LSA

0:18:59.440000 --> 0:19:00.340000
 as I've got in my database.

0:19:00.340000 --> 0:19:04.240000
 So in my database descriptor, I might
 say, hey, I have an OSPF LSA from

0:19:04.240000 --> 0:19:06.780000
 John with sequence number two.

0:19:06.780000 --> 0:19:11.280000
 I got an OSPF LSA from Sally with sequence
 number 17 and so on and so

0:19:11.280000 --> 0:19:14.360000
 forth. And then you'll give me your
 database descriptors and by the time

0:19:14.360000 --> 0:19:18.380000
 we're done exchanging database descriptors,
 we will know if we're not

0:19:18.380000 --> 0:19:21.140000
 synchronized, if we're missing
 something, something.

0:19:21.140000 --> 0:19:24.100000
 We probably are, we're probably
 not synchronized.

0:19:24.100000 --> 0:19:27.440000
 So then we're going to go into link
 state requests, link state updates

0:19:27.440000 --> 0:19:29.540000
 and link state acknowledgments.

0:19:29.540000 --> 0:19:32.700000
 So I will request from you, I'll say,
 hey, you gave me a database descriptor

0:19:32.700000 --> 0:19:37.240000
 and you said you had an LSA from Juan
 with a sequence number of 52.

0:19:37.240000 --> 0:19:38.580000
 I'm requesting that.

0:19:38.580000 --> 0:19:40.300000
 Can you please give that to me?

0:19:40.300000 --> 0:19:44.120000
 Then you'll give it to me in the
 form of a link state update.

0:19:44.120000 --> 0:19:48.260000
 So an OSPF link state update is the
 actual name of the OSPF packet type

0:19:48.260000 --> 0:19:52.320000
 that has LSAs inside of it.

0:19:52.320000 --> 0:19:55.860000
 People who are first learning about
 OSPF, a lot of times they make the

0:19:55.860000 --> 0:20:00.620000
 common mistake that they think that
 an OSPF LSA is the name of a packet

0:20:00.620000 --> 0:20:02.920000
 that, oh, we exchange LSAs.

0:20:02.920000 --> 0:20:08.640000
 No, we exchange link state updates and
 in the body of the link state update,

0:20:08.640000 --> 0:20:12.580000
 it has LSA information
 inside there.

0:20:12.580000 --> 0:20:16.500000
 For every link state update that I send
 you, you have to acknowledge it

0:20:16.500000 --> 0:20:18.800000
 and we use link state acknowledgments
 for that.

0:20:18.800000 --> 0:20:23.100000
 So OSPF does not use UDP.

0:20:23.100000 --> 0:20:24.760000
 It does not use TCP.

0:20:24.760000 --> 0:20:27.140000
 It is its own IP protocol.

0:20:27.140000 --> 0:20:30.300000
 So just like you could see that the
 IP header and then normally after

0:20:30.300000 --> 0:20:33.860000
 the IP header, you'd have a
 TCP header or a UDP header.

0:20:33.860000 --> 0:20:37.440000
 Well, if it's carrying OSPF, right after
 the IP header, you've got your

0:20:37.440000 --> 0:20:39.960000
 OSPF header and your OSPF data.

0:20:39.960000 --> 0:20:44.240000
 And yet because OSPF has link state acknowledgments,
 it has some connection

0:20:44.240000 --> 0:20:48.260000
 oriented properties similar
 to what TCP has.

0:20:48.260000 --> 0:20:53.140000
 So eventually, after we finish doing
 our link state updates and link state

0:20:53.140000 --> 0:20:57.280000
 acknowledgments, we'll reach this place
 of Nirvana where we say, hey,

0:20:57.280000 --> 0:20:58.700000
 we are synchronized.

0:20:58.700000 --> 0:21:02.540000
 You've got the exact same sets of LSAs
 that I have and now our neighbor

0:21:02.540000 --> 0:21:06.480000
 relationship will be complete.

0:21:06.480000 --> 0:21:12.500000
 Most OSPF packets are sent as multicasts
 and you need to know what multicasted

0:21:12.500000 --> 0:21:15.200000
 dresses are reserved for OSPF.

0:21:15.200000 --> 0:21:26.860000
 So in the world of IP version four,
 most OSPF packets go to 224.0.0.5.

0:21:26.860000 --> 0:21:31.620000
 Some of them go to 224.0.0.6.

0:21:31.620000 --> 0:21:36.480000
 So those two addresses, those class
 D multicast addresses are reserved

0:21:36.480000 --> 0:21:41.960000
 for OSPF. Now there is an equivalent
 of OSPF for IPV6.

0:21:41.960000 --> 0:21:48.580000
 For IPV4, it's actually
 OSPF version two.

0:21:48.580000 --> 0:21:52.800000
 For IPV6, it's OSPF version three.


0:21:52.800000 --> 0:21:55.980000
 OSPF version three uses FF.

0:21:55.980000 --> 0:21:59.400000
 Remember FF is multicast
 in the world of IPV6.

0:21:59.400000 --> 0:22:10.880000
 FF02 colon colon five
 and FF02 colon six.

0:22:10.880000 --> 0:22:11.020000
 So you can see that there are also some
 things that are in the world of

0:22:11.020000 --> 0:22:12.380000
 IPV6. So this is the one that you
 can see between the addresses.

0:22:12.380000 --> 0:22:21.820000
 Network types. All right, this is also
 something that's unique to OSPF.

0:22:21.820000 --> 0:22:27.480000
 So OSPF classifies links based
 on the layer two encapsulation.

0:22:27.480000 --> 0:22:32.920000
 So the moment you tell OSPF via a command,
 hey, this fast ethernet link,

0:22:32.920000 --> 0:22:34.420000
 I want you to start working on it.


0:22:34.420000 --> 0:22:36.880000
 This gigabit link, I want you
 to start working on it.

0:22:36.880000 --> 0:22:39.680000
 The first thing OSPF is going to do,
 it's going to look at the layer two

0:22:39.680000 --> 0:22:41.880000
 encapsulation on that link.

0:22:41.880000 --> 0:22:43.680000
 So it's going to say, oh, you
 want me to operate there.

0:22:43.680000 --> 0:22:44.460000
 All right, what are you doing?

0:22:44.460000 --> 0:22:50.660000
 Are you doing PPP, HDLC, Ethernet, frame
 relay, what have you got on that

0:22:50.660000 --> 0:22:52.740000
 interface as far as encapsulation?


0:22:52.740000 --> 0:22:57.560000
 And by getting the answer to that question,
 OSPF categorizes that link

0:22:57.560000 --> 0:23:01.640000
 as one of about four or five different
 network types, which we're going

0:23:01.640000 --> 0:23:03.280000
 to look at here in the next slide.


0:23:03.280000 --> 0:23:04.460000
 Now, why does it do that?

0:23:04.460000 --> 0:23:07.360000
 Why is it care what
 the network type is?

0:23:07.360000 --> 0:23:11.780000
 Well, because OSPF operates a little
 bit differently depending on what

0:23:11.780000 --> 0:23:14.440000
 network type it's connected to.

0:23:14.440000 --> 0:23:18.540000
 As it says things, as it says here, the
 network type will determine things

0:23:18.540000 --> 0:23:21.800000
 like, can I dynamically discover
 a neighbor across that link?

0:23:21.800000 --> 0:23:25.640000
 If the answer is no, based on that particular
 type of network, that means

0:23:25.640000 --> 0:23:27.820000
 you have to statically configure
 your neighbor.

0:23:27.820000 --> 0:23:35.820000
 You can see here, what's subnet mask
 is used to identify the link.

0:23:35.820000 --> 0:23:39.140000
 On some types of links, OSPF will just
 pick up whatever the subnet mask

0:23:39.140000 --> 0:23:41.020000
 is that you've configured
 on the interface.

0:23:41.020000 --> 0:23:47.160000
 If you configured 1111 slash 24, it'll
 advertise that link as slash 24.

0:23:47.160000 --> 0:23:51.600000
 But other types of links like loopback
 interfaces, it doesn't matter what

0:23:51.600000 --> 0:23:53.100000
 subnet mask you select.

0:23:53.100000 --> 0:23:55.840000
 It'll just always advertise
 that as a slash 32.

0:23:55.840000 --> 0:23:58.680000
 So that's another reason why it
 needs to identify the link type.

0:23:58.680000 --> 0:24:02.740000
 Is there a need for a designated router
 and a backup designated router

0:24:02.740000 --> 0:24:07.400000
 election? Some link types need
 that, some links types don't.

0:24:07.400000 --> 0:24:12.200000
 And the link type can be manually
 changed on most interface types.

0:24:12.200000 --> 0:24:18.580000
 So for example, if I go back to here,
 notice that when I enabled OSPF

0:24:18.580000 --> 0:24:22.600000
 on gig 01, it says, all right, based
 on the fact that this is running

0:24:22.600000 --> 0:24:27.300000
 ethernet, gigabit ethernet, I'm going
 to categorize this as a network

0:24:27.300000 --> 0:24:29.440000
 type of broadcast.

0:24:29.440000 --> 0:24:30.960000
 But I can change that.

0:24:30.960000 --> 0:24:37.600000
 I can tell OSPF, no, I want you to look
 at it a little bit differently.

0:24:37.600000 --> 0:24:42.560000
 IP OSPF network, I can choose
 from these right here.

0:24:42.560000 --> 0:24:48.140000
 Why don't you look at that as a
 point-to-point network instead?

0:24:48.140000 --> 0:24:51.220000
 And now he's no longer looking
 at the encapsulation type.

0:24:51.220000 --> 0:24:56.320000
 He's looking exactly what I told him
 and actually sees it as point-to

0:24:56.320000 --> 0:25:02.080000
-point. Now, at the CCNA level, you don't
 really need to know that command.

0:25:02.080000 --> 0:25:06.140000
 You don't really need to know why
 you would change the network type.

0:25:06.140000 --> 0:25:10.880000
 Just know that OSPF categorizes links
 based on their network type.

0:25:10.880000 --> 0:25:14.360000
 And that is something you can change
 via configuration should you want

0:25:14.360000 --> 0:25:24.560000
 to do that. At the CCNP level, you
 get formally learn that command.

0:25:24.560000 --> 0:25:28.880000
 All right, these are
 the network types.

0:25:28.880000 --> 0:25:33.520000
 And like I said, it's primarily based
 off the type of encapsulation that

0:25:33.520000 --> 0:25:39.120000
 is on there. So for the purposes of any
 Cisco exam or maybe other person,

0:25:39.120000 --> 0:25:43.280000
 other companies exam, you'll want to
 memorize what's on this slide right

0:25:43.280000 --> 0:25:49.120000
 here. Notice that point-to-multipoint
 is the only type of network that

0:25:49.120000 --> 0:25:52.620000
 OSPF cannot dynamically determine.


0:25:52.620000 --> 0:25:54.780000
 That's not based on any
 encapsulation type.

0:25:54.780000 --> 0:25:57.460000
 That's something you'd have
 to manually configure.

0:25:57.460000 --> 0:26:02.340000
 So once you, if you Google OSPF-point
-to-multipoint and you start learning

0:26:02.340000 --> 0:26:05.540000
 about the benefits of it, and you decide,
 oh, yeah, that's actually something

0:26:05.540000 --> 0:26:07.600000
 I could use. I want to do that.

0:26:07.600000 --> 0:26:11.760000
 That would require a configuration command
 on an interface to tell OSPF

0:26:11.760000 --> 0:26:14.040000
 that that's what you wanted to do.


0:26:14.040000 --> 0:26:17.600000
 And loopback, well, that's clearly
 for loopback interfaces.

0:26:17.600000 --> 0:26:22.980000
 All right, I mentioned that you and
 I exchanging hellos is the beginning

0:26:22.980000 --> 0:26:26.800000
 of the process. I said there are certain
 things within those hellos that

0:26:26.800000 --> 0:26:30.900000
 have to match between us if we're going
 to take the next step and exchange

0:26:30.900000 --> 0:26:33.220000
 those database descriptor packets.


0:26:33.220000 --> 0:26:34.720000
 What are those things?

0:26:34.720000 --> 0:26:36.260000
 Well, that's what we
 have right here.

0:26:36.260000 --> 0:26:38.780000
 So our timers have to be the same.


0:26:38.780000 --> 0:26:43.400000
 Like I said, usually those timers are
 going to be 10 seconds for the hello

0:26:43.400000 --> 0:26:46.640000
 and 40 seconds for
 the dead interval.

0:26:46.640000 --> 0:26:51.080000
 If I change those on my interface, whatever
 I change them to has to match

0:26:51.080000 --> 0:26:52.800000
 on your interface.

0:26:52.800000 --> 0:26:55.380000
 Otherwise, we will not
 become neighbors.

0:26:55.380000 --> 0:26:58.460000
 OSPF does have the ability
 to use authentication.

0:26:58.460000 --> 0:27:03.120000
 You know, this is a way to prevent rogue
 unauthorized routers from connecting

0:27:03.120000 --> 0:27:07.620000
 to your network and doing funky
 stuff with your routing tables.

0:27:07.620000 --> 0:27:09.320000
 Configure authentication.

0:27:09.320000 --> 0:27:13.040000
 Those rogue routers will not
 know what the password is.

0:27:13.040000 --> 0:27:14.860000
 And so we'll just ignore them.

0:27:14.860000 --> 0:27:19.140000
 So the authentication type and the authentication
 password have to match.

0:27:19.140000 --> 0:27:19.980000
 That's optional.

0:27:19.980000 --> 0:27:21.700000
 You don't have to do it.

0:27:21.700000 --> 0:27:23.520000
 I mentioned the area ID.

0:27:23.520000 --> 0:27:26.600000
 If you and I are directly connected
 across a link, we have to agree what

0:27:26.600000 --> 0:27:27.980000
 area this link is in.

0:27:27.980000 --> 0:27:30.640000
 The prefix length.

0:27:30.640000 --> 0:27:35.340000
 So what that's talking about is,
 you know, maybe here's me.

0:27:35.340000 --> 0:27:39.300000
 Here's you. We're connected
 across a link.

0:27:39.300000 --> 0:27:43.600000
 We both agree that this link
 is an area one, let's say.

0:27:43.600000 --> 0:27:48.380000
 I'm configured as 1.1.1.1.1.1.1.2.


0:27:48.380000 --> 0:27:51.100000
 But somebody screwed up.

0:27:51.100000 --> 0:27:54.860000
 And on my side, I'm configured
 with this subnet mask.

0:27:54.860000 --> 0:27:59.800000
 On your side, you're configured
 with this subnet mask.

0:27:59.800000 --> 0:28:02.540000
 That is actually carried
 inside the hello packet.

0:28:02.540000 --> 0:28:06.280000
 And we will see that our subnet masks,
 otherwise known here as prefix

0:28:06.280000 --> 0:28:08.480000
 length, are not the same.

0:28:08.480000 --> 0:28:11.680000
 And if they're not the same,
 we can't be neighbors.

0:28:11.680000 --> 0:28:13.940000
 That's clearly a configuration
 error, though.

0:28:13.940000 --> 0:28:16.260000
 You'd want to fix that.

0:28:16.260000 --> 0:28:18.720000
 And the stub area flag.

0:28:18.720000 --> 0:28:22.520000
 At the CCI level, you don't really
 learn what stub areas are.

0:28:22.520000 --> 0:28:27.500000
 Stub areas are special types of areas
 that say certain types of LSA's

0:28:27.500000 --> 0:28:29.120000
 are not allowed in here.

0:28:29.120000 --> 0:28:32.960000
 We're going to allow some LSAs, but other
 special types of LSAs keep those

0:28:32.960000 --> 0:28:37.520000
 out. So if you configure an area as
 a stub area, all the routers in that

0:28:37.520000 --> 0:28:40.860000
 area have to know that, and
 they have to agree to that.

0:28:40.860000 --> 0:28:47.740000
 All right, so we've already
 talked about this.

0:28:47.740000 --> 0:28:49.980000
 There's nothing new here.

0:28:49.980000 --> 0:28:57.000000
 What this is referring to is, you know,
 once again, like I just said a

0:28:57.000000 --> 0:29:02.340000
 couple minutes ago, once I enable OSPF
 on myself on at least one interface,

0:29:02.340000 --> 0:29:06.800000
 I'm going to create a link state advertisement
 describing myself and that

0:29:06.800000 --> 0:29:10.060000
 interface and any neighbors
 I have on the interface.

0:29:10.060000 --> 0:29:12.540000
 All right, I own that LSA.

0:29:12.540000 --> 0:29:13.720000
 I send it out there.

0:29:13.720000 --> 0:29:17.600000
 It gets flooded to all the routers
 in the same area as me.

0:29:17.600000 --> 0:29:20.340000
 That LSA is not good forever.

0:29:20.340000 --> 0:29:21.700000
 It has a lifetime.

0:29:21.700000 --> 0:29:26.260000
 The lifetime of LSAs by default
 is one hour, 60 minutes.

0:29:26.260000 --> 0:29:30.680000
 So if you never see that LSA from me
 again, after 60 minutes, you will

0:29:30.680000 --> 0:29:35.040000
 purge it. And whatever routes were
 generated as a result of that LSA,

0:29:35.040000 --> 0:29:37.500000
 you'll strip those routes
 out of your routing table.

0:29:37.500000 --> 0:29:42.460000
 So the originator of the LSA, which is
 me, I'm going to refresh that every

0:29:42.460000 --> 0:29:45.660000
 30 minutes. So even if nothing changes
 and nothing's happened with this

0:29:45.660000 --> 0:29:49.360000
 link every 30 minutes, I will increment
 my sequence number in that LSA

0:29:49.360000 --> 0:29:56.840000
 by one, and I will send it
 out and flood it again.

0:29:56.840000 --> 0:30:00.000000
 We already talked about area
 zero being the backbone.

0:30:00.000000 --> 0:30:05.000000
 We already talked about what an area
 border router is, just as a review.

0:30:05.000000 --> 0:30:09.680000
 He's got one, at least one, link connected
 to area zero, and at least

0:30:09.680000 --> 0:30:14.280000
 one link connected to any
 other non-zero area.

0:30:14.280000 --> 0:30:20.760000
 And if I have a router that is doing
 something other than OSPF, like maybe

0:30:20.760000 --> 0:30:27.020000
 he's doing RIP or EIGR Pities, learned
 about a bunch of non-OSPF routes,

0:30:27.020000 --> 0:30:31.780000
 and then he wants to convert those into
 OSPF LSA's and inject them into

0:30:31.780000 --> 0:30:33.440000
 the OSPF domain.

0:30:33.440000 --> 0:30:35.960000
 We call that an autonomous
 system border router.

0:30:35.960000 --> 0:30:39.720000
 Depending on the paper or document you
 read, sometimes they'll call that

0:30:39.720000 --> 0:30:44.420000
 an autonomous system boundary router,
 but everybody knows what ASBR means,

0:30:44.420000 --> 0:30:45.820000
 whether it's border or boundary.

0:30:45.820000 --> 0:30:51.520000
 It's a router that's injecting
 non-OSPF stuff into OSPF.

0:30:51.520000 --> 0:30:56.960000
 All right, there's
 nothing new here.

0:30:56.960000 --> 0:30:59.980000
 We've already talked about what link
 state advertisements are that they

0:30:59.980000 --> 0:31:03.000000
 age out after one hour.

0:31:03.000000 --> 0:31:09.580000
 OSPF fully supports variable links
 subnet masking and classless inter

0:31:09.580000 --> 0:31:16.960000
-domain routing. It does support summarization,
 but at the CCI level, you're

0:31:16.960000 --> 0:31:20.300000
 not expected to know how to do summarization
 or how to troubleshoot it.

0:31:20.300000 --> 0:31:24.120000
 If you're not familiar with that term
 summarization, all that means is,

0:31:24.120000 --> 0:31:29.140000
 imagine that you and I, once again,
 are neighbors, you and I are in area

0:31:29.140000 --> 0:31:34.000000
 one. Now, imagine that behind me, I have
 another link to another neighbor

0:31:34.000000 --> 0:31:37.480000
 who's I am an area border router.

0:31:37.480000 --> 0:31:39.340000
 I'm straddling two areas.

0:31:39.340000 --> 0:31:44.960000
 Now, imagine area zero, I learned about
 200 different routes that all

0:31:44.960000 --> 0:31:52.560000
 start with the number 20, 20.1.1 slash
 24, 20.56.90, all sorts of routes

0:31:52.560000 --> 0:31:53.720000
 that begin with 20.

0:31:53.720000 --> 0:31:58.420000
 Well, the normal behavior is, I would
 send you all those hundreds of routes.

0:31:58.420000 --> 0:32:02.280000
 Well, summarization means,
 hey, I'm your only way out.

0:32:02.280000 --> 0:32:05.260000
 You can't get to those networks
 except through me.

0:32:05.260000 --> 0:32:07.140000
 They all start with the number 20.


0:32:07.140000 --> 0:32:11.060000
 So, while I just summarize them down
 and just send you one LSA saying,

0:32:11.060000 --> 0:32:15.940000
 hey, any route, any packet you get
 that needs to go to 20.anything, go

0:32:15.940000 --> 0:32:19.640000
 through me. So, I'm not going to tell
 you about those specific routes.

0:32:19.640000 --> 0:32:21.400000
 They all have something in common.


0:32:21.400000 --> 0:32:23.220000
 I'll tell you about what
 they have in common.

0:32:23.220000 --> 0:32:27.120000
 That's route summarization.

0:32:27.120000 --> 0:32:33.300000
 We already talked about how 224, 005
 and 006 are reserved for OSPF.

0:32:33.300000 --> 0:32:36.340000
 And we'll talk about designated routers
 and backup designated routers

0:32:36.340000 --> 0:32:38.560000
 in a subsequent video.

0:32:38.560000 --> 0:32:42.980000
 And I talked about how OSPF supports
 special area types, such as stub,

0:32:42.980000 --> 0:32:45.180000
 totally stub, and not
 so stubby area.

0:32:45.180000 --> 0:32:49.040000
 As I mentioned, the main thing you just
 need to know about those is that,

0:32:49.040000 --> 0:32:52.640000
 well, I'll just give you a real high
 level, real high level over you of

0:32:52.640000 --> 0:32:59.740000
 that. So, imagine that we
 have something like this.

0:32:59.740000 --> 0:33:03.480000
 All right. This is area zero.

0:33:03.480000 --> 0:33:05.140000
 This is area 15.

0:33:05.140000 --> 0:33:07.160000
 And this is area 27.

0:33:07.160000 --> 0:33:11.800000
 All right. And we have
 a couple of ABRs.

0:33:11.800000 --> 0:33:18.540000
 Here's ABR2. Here's ABR1.

0:33:18.540000 --> 0:33:27.020000
 Okay. And then on area 15, right
 here, we have an ASBR.

0:33:27.020000 --> 0:33:35.660000
 So, this guy is learning some non-OSPF
 routes, maybe RIP, maybe EIGRP.

0:33:35.660000 --> 0:33:37.280000
 Who knows what they are.

0:33:37.280000 --> 0:33:41.740000
 And what he's doing is he's
 injecting those into OSPF.

0:33:41.740000 --> 0:33:46.680000
 Now, he's going to use a special type
 of LSA, not the type one router

0:33:46.680000 --> 0:33:50.540000
 LSA. He's going to use a different
 type of LSA to inject those.

0:33:50.540000 --> 0:33:55.900000
 And one thing that's very unique about
 that LSA is that LSA actually floods

0:33:55.900000 --> 0:33:58.660000
 throughout the entire OSPF domain.


0:33:58.660000 --> 0:34:01.960000
 So, that would come in
 here in the area zero.

0:34:01.960000 --> 0:34:05.120000
 And it would come in
 here in the area 27.

0:34:05.120000 --> 0:34:10.280000
 Well, sometimes you might say, hey,
 you know what, in area 27, ABR2 is

0:34:10.280000 --> 0:34:12.520000
 our only way in and out.

0:34:12.520000 --> 0:34:17.440000
 That's it. So, in area 27, we
 don't need all that stuff.

0:34:17.440000 --> 0:34:20.020000
 As a matter of fact, why don't we just,
 you know, it would really reduce

0:34:20.020000 --> 0:34:23.540000
 the size of our memory, reduce the
 size of our routing tables and our

0:34:23.540000 --> 0:34:28.140000
 database if instead of getting those
 routes, which, you know, we need

0:34:28.140000 --> 0:34:29.620000
 reachability to that.

0:34:29.620000 --> 0:34:31.380000
 But what if we did this instead?

0:34:31.380000 --> 0:34:37.560000
 What if we just told this guy to send
 one LSA that had a default route?

0:34:37.560000 --> 0:34:40.040000
 And not to send all
 that other stuff.

0:34:40.040000 --> 0:34:42.640000
 That's called a stub area.

0:34:42.640000 --> 0:34:45.760000
 Now, it gets more complicated than that,
 but that's the essence of a stub

0:34:45.760000 --> 0:34:50.400000
 area that says, hey, keep those redistributed
 routes out of here.

0:34:50.400000 --> 0:34:54.240000
 I don't want to know those, and I'll
 just use a single default route if

0:34:54.240000 --> 0:34:57.740000
 I need to send packets
 to an unknown network.

0:34:57.740000 --> 0:34:59.600000
 That's a stub area.

0:34:59.600000 --> 0:35:08.660000
 All right. And the last thing I want
 to review here are the various tables

0:35:08.660000 --> 0:35:11.620000
 and data structures
 that OSPF uses.

0:35:11.620000 --> 0:35:17.300000
 You should be aware that OSPF maintains
 or feeds into all of these.

0:35:17.300000 --> 0:35:20.620000
 So OSPF maintains
 a neighbor table.

0:35:20.620000 --> 0:35:33.540000
 And you can view that with show IP OSPF
 neighbor to see if you have any

0:35:33.540000 --> 0:35:37.560000
 OSPF neighbors. Those are
 maintained in the table.

0:35:37.560000 --> 0:35:42.840000
 All the LSA is that you've either created
 yourself or LSA's that you've

0:35:42.840000 --> 0:35:47.740000
 learned from other people are stored
 in your link state database.

0:35:47.740000 --> 0:35:56.740000
 And you can view the contents of
 that with show IP OSPF database.

0:35:56.740000 --> 0:36:04.740000
 Database. Forget my, forgive my
 sloppy writing right there.

0:36:04.740000 --> 0:36:10.500000
 The SPF tree is like your jigsaw puzzle
 where you finally put together

0:36:10.500000 --> 0:36:14.120000
 everything into a coherent
 map, coherent picture.

0:36:14.120000 --> 0:36:17.300000
 I have searched and I have
 not been able to find.

0:36:17.300000 --> 0:36:18.680000
 I don't know if it exists.

0:36:18.680000 --> 0:36:22.320000
 I don't know of any commands that you
 can issue to actually see the SPF

0:36:22.320000 --> 0:36:25.920000
 tree. I think that's just a bunch of
 memory cells linked together and

0:36:25.920000 --> 0:36:29.720000
 dram memory or something that the
 CPU of the router can access.

0:36:29.720000 --> 0:36:33.340000
 But I don't think we have visibility
 to that ourselves.

0:36:33.340000 --> 0:36:38.020000
 And then, of course, the routes
 are extracted from the SPF tree.

0:36:38.020000 --> 0:36:40.920000
 And the best routes are put
 into the IP routing table.

0:36:40.920000 --> 0:36:45.940000
 And we can certainly
 do show IP route.

0:36:45.940000 --> 0:36:50.460000
 But if you just want to look at OSPF
 routes, not show IP routes, show

0:36:50.460000 --> 0:36:58.300000
 IP route, you can tack onto that show
 IP route OSPF to just view the OSPF

0:36:58.300000 --> 0:37:08.120000
 learned routes. So that concludes
 this video refresher on OSPF.

0:37:08.120000 --> 0:37:09.380000
 I hope you enjoyed it.

