1
00:00:00,000 --> 00:00:07,080

2
00:00:07,080 --> 00:00:09,540
So now that we've seen
some of the problems

3
00:00:09,540 --> 00:00:12,540
with not doing
summarization, let's

4
00:00:12,540 --> 00:00:17,010
begin by actually looking at
summarizing internal routes,

5
00:00:17,010 --> 00:00:21,460
or type III LSAs, in OSPF.

6
00:00:21,460 --> 00:00:23,860
Internal and external
summarization

7
00:00:23,860 --> 00:00:25,450
are done differently.

8
00:00:25,450 --> 00:00:28,850
Now, they actually
function very similar.

9
00:00:28,850 --> 00:00:31,510
However, there are
different restrictions

10
00:00:31,510 --> 00:00:35,020
on when we can do each
one, and the command

11
00:00:35,020 --> 00:00:38,680
itself is different for
internal versus external.

12
00:00:38,680 --> 00:00:42,388
So as usual, let's take a look
at our diagram real quick,

13
00:00:42,388 --> 00:00:43,930
so we can be familiar
with what we're

14
00:00:43,930 --> 00:00:46,190
going to be doing in
the demonstration.

15
00:00:46,190 --> 00:00:49,960
And then we'll jump into the CLI
and see how summarization works

16
00:00:49,960 --> 00:00:54,000
for our internal prefixes.

17
00:00:54,000 --> 00:00:56,900
So in this video,
we'll be focusing just

18
00:00:56,900 --> 00:00:58,620
on the internal routes.

19
00:00:58,620 --> 00:01:00,980
So since we're focusing
on the internals,

20
00:01:00,980 --> 00:01:03,770
we'll just be looking
at those four routes

21
00:01:03,770 --> 00:01:07,220
that we saw earlier,
coming from router 10.

22
00:01:07,220 --> 00:01:10,250
These are, of course,
coming down to the ABR.

23
00:01:10,250 --> 00:01:14,000
In this case, that's router
6 is acting as the ABR

24
00:01:14,000 --> 00:01:15,590
here for area 200.

25
00:01:15,590 --> 00:01:22,100
Remember that in OSPF, you
can only summarize on an ABR.

26
00:01:22,100 --> 00:01:25,250
In this case, that would
be router 6, router 1,

27
00:01:25,250 --> 00:01:26,760
or router 3.

28
00:01:26,760 --> 00:01:30,330
Those are the only approaches
we have in this topology.

29
00:01:30,330 --> 00:01:32,030
So what we're going
to do is, we're

30
00:01:32,030 --> 00:01:37,110
going to take these 4/24
routes that we have right now,

31
00:01:37,110 --> 00:01:43,550
and we are going to summarize
them into 1/16 route being

32
00:01:43,550 --> 00:01:46,220
advertised by router 6.

33
00:01:46,220 --> 00:01:48,230
Then we'll, of
course, take a look

34
00:01:48,230 --> 00:01:51,560
at the changes that
creates on router 7

35
00:01:51,560 --> 00:01:56,660
as far as, once again, causing
a failure on router 10,

36
00:01:56,660 --> 00:01:59,060
just for one of the
loopbacks, and then

37
00:01:59,060 --> 00:02:01,680
bringing that interface back up.

38
00:02:01,680 --> 00:02:05,510
So we'll take a look at what
happens when that route flaps.

39
00:02:05,510 --> 00:02:08,030
So let's jump into
the command line.

40
00:02:08,030 --> 00:02:09,979
As usual, we'll
start with looking

41
00:02:09,979 --> 00:02:11,390
at what we already have.

42
00:02:11,390 --> 00:02:15,380
So we'll start on router 7 and
look at these existing routes,

43
00:02:15,380 --> 00:02:17,420
just to verify that
we're getting them

44
00:02:17,420 --> 00:02:20,330
before we do our summarization.

45
00:02:20,330 --> 00:02:23,140
So we'll start with
the show IP route OSPF.

46
00:02:23,140 --> 00:02:25,690
We'll look specifically at
the anterior area routes.

47
00:02:25,690 --> 00:02:29,590
And what we're looking for
are these 100.100 networks.

48
00:02:29,590 --> 00:02:34,750
We can see that right now, we do
have 4/24 networks, of course,

49
00:02:34,750 --> 00:02:37,700
only being different
in the third octet.

50
00:02:37,700 --> 00:02:45,280
So we're going to be summarizing
these to 100.100.0.0/16.

51
00:02:45,280 --> 00:02:47,680
To do this, of course,
we have to go to our ABR.

52
00:02:47,680 --> 00:02:50,320
So let's jump up to router 6.

53
00:02:50,320 --> 00:02:52,150
We'll go into
configuration mode.

54
00:02:52,150 --> 00:02:54,700
Of course, into
our OSPF process.

55
00:02:54,700 --> 00:02:57,320
And then we're going to
do the area range command.

56
00:02:57,320 --> 00:03:00,040
Now when we put
in the area, this

57
00:03:00,040 --> 00:03:02,990
is going to be the area
that the routes are in.

58
00:03:02,990 --> 00:03:05,710
So in this case, it's area 200.

59
00:03:05,710 --> 00:03:10,610
We will be summarizing
into all other areas.

60
00:03:10,610 --> 00:03:13,240
So just keep in mind, when
you're doing the area range

61
00:03:13,240 --> 00:03:16,840
command, it's always the source
area that you're listing,

62
00:03:16,840 --> 00:03:20,200
the one that has the prefixes
that we wish to summarize.

63
00:03:20,200 --> 00:03:23,240
Out of area 200, of course,
we're doing a range.

64
00:03:23,240 --> 00:03:25,240
This is why we call it
the area range command.

65
00:03:25,240 --> 00:03:27,970
And then we put in the
network that we actually

66
00:03:27,970 --> 00:03:29,300
want to summarize.

67
00:03:29,300 --> 00:03:33,070
So in this case, it's
going to be the 100.100.

68
00:03:33,070 --> 00:03:35,330
And of course, then
we put in the mask.

69
00:03:35,330 --> 00:03:37,120
We're going to make this a /16.

70
00:03:37,120 --> 00:03:40,340
There's a few other options
we can do on this as well.

71
00:03:40,340 --> 00:03:42,800
So for example, we
could set the cost.

72
00:03:42,800 --> 00:03:46,940
Notice the default is
to advertise this route.

73
00:03:46,940 --> 00:03:49,800
We're not going to change any
of these options right now.

74
00:03:49,800 --> 00:03:53,270
So let's just go
ahead and hit Enter.

75
00:03:53,270 --> 00:03:56,690
And let's do a show IP route.

76
00:03:56,690 --> 00:04:00,680
The reason we're doing a
show IP route on this device

77
00:04:00,680 --> 00:04:03,140
is, we want to see that
it does still, of course,

78
00:04:03,140 --> 00:04:06,180
itself have those 4/24s.

79
00:04:06,180 --> 00:04:10,190
But notice that
we now have a /16.

80
00:04:10,190 --> 00:04:16,519
It's listed as a summary, and
the next hop is null zero.

81
00:04:16,519 --> 00:04:20,810
This route is referred
to as the discard route.

82
00:04:20,810 --> 00:04:27,050
Its purpose is to prevent
loops in our OSPF environment.

83
00:04:27,050 --> 00:04:33,000
Say, for example, we had an
advertisement of this summary.

84
00:04:33,000 --> 00:04:36,900
And since we're saying that
we can get to all of 100.100,

85
00:04:36,900 --> 00:04:42,180
well what if somebody sent
a packet to us going to, say

86
00:04:42,180 --> 00:04:44,080
a network that doesn't exist?

87
00:04:44,080 --> 00:04:50,460
So maybe it's going
to 100.100.200.0/24.

88
00:04:50,460 --> 00:04:55,230
Well, we don't have a
network that's 200/24.

89
00:04:55,230 --> 00:04:59,340
Now normally, what we would do
is just discard that traffic,

90
00:04:59,340 --> 00:05:01,960
because we don't
have a route for it.

91
00:05:01,960 --> 00:05:08,200
But what if this router happens
to also have a default route?

92
00:05:08,200 --> 00:05:11,130
So maybe this is like
an internet edge,

93
00:05:11,130 --> 00:05:14,970
and we're default
routing to the next hop,

94
00:05:14,970 --> 00:05:18,060
and we're advertising
this summary.

95
00:05:18,060 --> 00:05:21,100
Then somebody sends us
something that doesn't exist.

96
00:05:21,100 --> 00:05:22,600
We don't have a route for it.

97
00:05:22,600 --> 00:05:25,500
So we send it right back to the
default gateway, who of course

98
00:05:25,500 --> 00:05:26,750
routes it right back to us.

99
00:05:26,750 --> 00:05:28,780
We send it right back
to the default gateway.

100
00:05:28,780 --> 00:05:31,620
And just like that,
we have a loop.

101
00:05:31,620 --> 00:05:36,210
By putting in this summary
route to null zero,

102
00:05:36,210 --> 00:05:39,900
this is covering the entire
range we're advertising.

103
00:05:39,900 --> 00:05:42,210
And essentially,
what we're saying is,

104
00:05:42,210 --> 00:05:47,940
if any traffic shows up
for 100.100 dot anything,

105
00:05:47,940 --> 00:05:52,260
if we don't have a more specific
route to actually deliver it,

106
00:05:52,260 --> 00:05:57,270
like these 4/24s that we have
right now, just discard it.

107
00:05:57,270 --> 00:05:59,820
This way, we're not going
to try to send it back

108
00:05:59,820 --> 00:06:01,840
over something like a default.

109
00:06:01,840 --> 00:06:05,070
So again, this is critical
for loop prevention.

110
00:06:05,070 --> 00:06:08,730
And as we can see, OSPF
puts this discard route

111
00:06:08,730 --> 00:06:12,360
in the routing table for
us automatically on the ABR

112
00:06:12,360 --> 00:06:14,470
doing the summarization.

113
00:06:14,470 --> 00:06:16,650
So now let's go back
over to router 7

114
00:06:16,650 --> 00:06:20,160
and see what this looks like
on a router that's not the ABR.

115
00:06:20,160 --> 00:06:23,970
So on router 7, we'll simply
repeat the last command again.

116
00:06:23,970 --> 00:06:32,250
But this time, notice that we
have a 100.100.0.0/16 network.

117
00:06:32,250 --> 00:06:33,940
That is, of course, our summary.

118
00:06:33,940 --> 00:06:36,390
So now let's take a look at
the effects this is going

119
00:06:36,390 --> 00:06:38,100
to have on our convergence.

120
00:06:38,100 --> 00:06:41,280
Remember, we saw previously
that if a network flaps

121
00:06:41,280 --> 00:06:45,660
on router 10, router 7
doesn't have to run Dijkstra.

122
00:06:45,660 --> 00:06:48,640
But it does still have to
update the routing table.

123
00:06:48,640 --> 00:06:51,240
So we still saw
updates coming in.

124
00:06:51,240 --> 00:06:53,190
Let's go ahead and
turn on our debug,

125
00:06:53,190 --> 00:06:57,340
and let's go flap a
network on router 10.

126
00:06:57,340 --> 00:06:59,790
Let's just pick on loopback 110.

127
00:06:59,790 --> 00:07:03,180
We'll shut this interface and
jump back over to router 7.

128
00:07:03,180 --> 00:07:07,800
Notice that absolutely nothing
happened here in our debug.

129
00:07:07,800 --> 00:07:11,070
This is the summary
route, protecting us

130
00:07:11,070 --> 00:07:15,270
from root flaps in the other
area behind the summary.

131
00:07:15,270 --> 00:07:17,920
And this is one of
our main purposes.

132
00:07:17,920 --> 00:07:20,730
Now something else to
note while we're here--

133
00:07:20,730 --> 00:07:24,360
if we look back up at our
routing table for this summary,

134
00:07:24,360 --> 00:07:28,200
notice that the metric on
that summary is currently 3.

135
00:07:28,200 --> 00:07:30,360
Because the next
thing we want to look

136
00:07:30,360 --> 00:07:36,610
at is how OSPF calculates
the metric for the summary.

137
00:07:36,610 --> 00:07:39,480
So to do this, let's go
back over to router 10.

138
00:07:39,480 --> 00:07:41,880
First thing we'll do is
no shut the interface.

139
00:07:41,880 --> 00:07:45,630
Then let's change the OSPF cost
on some of these interfaces.

140
00:07:45,630 --> 00:07:48,010
Since we're already
under interface 110,

141
00:07:48,010 --> 00:07:51,720
let's change this one
to say a cost of 20.

142
00:07:51,720 --> 00:07:54,090
We'll go in interface
loopback 100.

143
00:07:54,090 --> 00:07:56,160
We'll change this
one to a cost of 10.

144
00:07:56,160 --> 00:07:58,470
And then let's go
into loopback 120.

145
00:07:58,470 --> 00:08:00,850
We'll make this one
to cost of 20 as well.

146
00:08:00,850 --> 00:08:03,480
And we also have a loopback 130.

147
00:08:03,480 --> 00:08:05,440
We'll make that
one a cost of 30.

148
00:08:05,440 --> 00:08:08,530
So these are the loopbacks that
represent each of our networks.

149
00:08:08,530 --> 00:08:10,280
So what we're doing
is, we're setting them

150
00:08:10,280 --> 00:08:13,180
all to different costs so
we can see what happens.

151
00:08:13,180 --> 00:08:15,060
Let's go back over to router 7.

152
00:08:15,060 --> 00:08:17,100
Now the first thing to
note here on router 7

153
00:08:17,100 --> 00:08:21,030
is that we have, in fact,
gotten SPF updates now.

154
00:08:21,030 --> 00:08:24,880
This is because the metric on
that route actually changed.

155
00:08:24,880 --> 00:08:26,790
Let's take a look at
our routing table again.

156
00:08:26,790 --> 00:08:31,300
Notice that the metric
on that route is now 12.

157
00:08:31,300 --> 00:08:34,559
That means that it's
actually using the loopback

158
00:08:34,559 --> 00:08:39,730
with a cost of 10, plus there's
two links between us and there.

159
00:08:39,730 --> 00:08:42,280
So it's adding 2 to that.

160
00:08:42,280 --> 00:08:44,520
But to summarize
what it's doing here,

161
00:08:44,520 --> 00:08:47,250
it's actually using
the lowest metric

162
00:08:47,250 --> 00:08:49,680
from the routes that
are being summarized.

163
00:08:49,680 --> 00:08:53,100
So of course, we set some to
10, some to 20, some to 30,

164
00:08:53,100 --> 00:08:54,690
and it's using the 10.

165
00:08:54,690 --> 00:08:58,750
It's, of course, also counting
the other interfaces involved.

166
00:08:58,750 --> 00:09:01,800
But overall, it's
using the lowest cost

167
00:09:01,800 --> 00:09:03,870
of the routes being summarized.

168
00:09:03,870 --> 00:09:06,840
We can, of course, verify this
by going back over the router

169
00:09:06,840 --> 00:09:09,360
10, going into our
interface loopback

170
00:09:09,360 --> 00:09:12,490
100, which is the one
with the lowest cost,

171
00:09:12,490 --> 00:09:14,940
therefore the one that it
would currently be using,

172
00:09:14,940 --> 00:09:18,480
and shut this network down.

173
00:09:18,480 --> 00:09:21,610
Now here's sort of a downside
that we have to be aware of.

174
00:09:21,610 --> 00:09:24,600
If we come back over
here to router 7, notice,

175
00:09:24,600 --> 00:09:26,730
we've gotten another
update again.

176
00:09:26,730 --> 00:09:28,750
Now we expected this behavior.

177
00:09:28,750 --> 00:09:32,250
Because if we look at the
route now, now it's 22.

178
00:09:32,250 --> 00:09:36,670
So obviously, it has to send an
update for that metric change.

179
00:09:36,670 --> 00:09:38,880
But herein lies the problem.

180
00:09:38,880 --> 00:09:42,030
The whole goal, or at least
one of the major goals,

181
00:09:42,030 --> 00:09:46,320
of doing this summarization
was to prevent this.

182
00:09:46,320 --> 00:09:50,190
We didn't want the
device to receive updates

183
00:09:50,190 --> 00:09:52,080
every time a route flaps.

184
00:09:52,080 --> 00:09:54,600
And if they have
different metrics,

185
00:09:54,600 --> 00:09:57,220
that may very well still happen.

186
00:09:57,220 --> 00:10:00,030
So one thing we might
want to do to fix this

187
00:10:00,030 --> 00:10:04,950
is to actually hard code
the metric on the summary.

188
00:10:04,950 --> 00:10:08,980
That way, if the routes behind
the summary, particularly,

189
00:10:08,980 --> 00:10:12,300
of course, the one with the
lowest metric, were to fail,

190
00:10:12,300 --> 00:10:14,250
we wouldn't still
get this update

191
00:10:14,250 --> 00:10:15,900
to we're trying to prevent.

192
00:10:15,900 --> 00:10:18,540
So that might be a
good practice that you

193
00:10:18,540 --> 00:10:20,880
want to do in the real world.

194
00:10:20,880 --> 00:10:22,980
To do that, of course,
all we have to do

195
00:10:22,980 --> 00:10:29,030
is go back over to router 6, go
back to our area range command,

196
00:10:29,030 --> 00:10:31,160
and simply set the cost.

197
00:10:31,160 --> 00:10:32,240
We'll set it to 100.

198
00:10:32,240 --> 00:10:35,000
Now this will, of course,
cause one more update

199
00:10:35,000 --> 00:10:37,280
to go into area 0.

200
00:10:37,280 --> 00:10:39,440
So if we go back
over to router 7,

201
00:10:39,440 --> 00:10:42,380
we see that once again,
we've gotten another update.

202
00:10:42,380 --> 00:10:45,020
But hopefully, this should
then be the last one,

203
00:10:45,020 --> 00:10:46,950
regardless of what happens.

204
00:10:46,950 --> 00:10:48,770
So again, if we look
at the route now,

205
00:10:48,770 --> 00:10:54,530
we see that it just flat out has
a metric of 101, the 100 that's

206
00:10:54,530 --> 00:10:59,820
being advertised plus our cost
of 1 to get to router 6 itself.

207
00:10:59,820 --> 00:11:02,390
Now if those networks
on router 10 flap,

208
00:11:02,390 --> 00:11:06,320
it won't change anything,
even if they're a better cost.

209
00:11:06,320 --> 00:11:09,530
Let's bring that interface
back up on router 10.

210
00:11:09,530 --> 00:11:12,150
Remember that this is the
one with the best cost.

211
00:11:12,150 --> 00:11:14,810
So theoretically, based
on the normal rules,

212
00:11:14,810 --> 00:11:16,700
this should now
become the metric.

213
00:11:16,700 --> 00:11:18,950
But of course, we
know that it will not.

214
00:11:18,950 --> 00:11:22,040
We see here on router 7 that
we did not get an update.

215
00:11:22,040 --> 00:11:24,950
Everything has been
quiet, even though we had

216
00:11:24,950 --> 00:11:26,750
a better route come back up.

217
00:11:26,750 --> 00:11:31,340
Now of course, it's still
going to have that cost of 101.

218
00:11:31,340 --> 00:11:34,520
This isn't going to change
now, regardless of what

219
00:11:34,520 --> 00:11:37,440
those routes behind it do.

220
00:11:37,440 --> 00:11:40,440
In this video, we
took a look at how

221
00:11:40,440 --> 00:11:44,610
to summarize internal
routes in OSPF.

222
00:11:44,610 --> 00:11:46,740
Some of our key
takeaways-- remember,

223
00:11:46,740 --> 00:11:49,800
we can only do this on the ABR.

224
00:11:49,800 --> 00:11:54,090
It uses the area range
command to accomplish this.

225
00:11:54,090 --> 00:11:57,840
And by default, it
will use the metric

226
00:11:57,840 --> 00:12:00,690
from the lowest cost
route that's being

227
00:12:00,690 --> 00:12:03,930
summarized for the summary.

228
00:12:03,930 --> 00:12:06,600
But to help increase
stability, we

229
00:12:06,600 --> 00:12:08,490
may want to go ahead
and take care of that

230
00:12:08,490 --> 00:12:13,410
ourselves by manually setting
a metric on that summary.

231
00:12:13,410 --> 00:12:15,540
I hope this has been
informative for you,

232
00:12:15,540 --> 00:12:18,530
and I'd like to thank
you you for viewing.

233
00:12:18,530 --> 00:12:22,000