1
00:00:00,000 --> 00:00:00,500

2
00:00:00,500 --> 00:00:07,000
[AUDIO LOGO]

3
00:00:07,000 --> 00:00:09,970
So let's begin by
jumping in and taking

4
00:00:09,970 --> 00:00:13,990
a look at the reasons
or advantages of doing

5
00:00:13,990 --> 00:00:15,530
route summarization.

6
00:00:15,530 --> 00:00:17,860
Now, the main things we're
going to be looking at

7
00:00:17,860 --> 00:00:19,810
and there are many
different advantages.

8
00:00:19,810 --> 00:00:23,020
But the two main focus
is first, reducing

9
00:00:23,020 --> 00:00:25,060
the size of the routing table.

10
00:00:25,060 --> 00:00:28,120
Obviously, this is going to
help increase performance

11
00:00:28,120 --> 00:00:30,700
and help with scalability.

12
00:00:30,700 --> 00:00:33,010
Our network can get
larger and larger as long

13
00:00:33,010 --> 00:00:35,020
as we're summarizing properly.

14
00:00:35,020 --> 00:00:38,170
This is a big deal
particularly with OSPF

15
00:00:38,170 --> 00:00:41,620
since all of the routers
have to maintain a database.

16
00:00:41,620 --> 00:00:46,220
The other main thing we want
to focus on is fault isolation.

17
00:00:46,220 --> 00:00:49,570
So if a route fails
in another area,

18
00:00:49,570 --> 00:00:52,480
should all of the
areas really have

19
00:00:52,480 --> 00:00:56,470
to recalculate their database
and update their routing table

20
00:00:56,470 --> 00:00:58,850
because of some remote failure.

21
00:00:58,850 --> 00:01:04,540
Now, keep in mind we can only
do summarization between areas

22
00:01:04,540 --> 00:01:09,100
or for external routes as
they're brought into OSPF.

23
00:01:09,100 --> 00:01:11,800
Now, that being the
case, we would not

24
00:01:11,800 --> 00:01:15,890
be talking about type
I or type 2 LSAs here.

25
00:01:15,890 --> 00:01:18,910
So we would not be
rerunning Dijkstra.

26
00:01:18,910 --> 00:01:23,260
So this is not about saving
us having to run the SPF

27
00:01:23,260 --> 00:01:25,850
algorithm within an area.

28
00:01:25,850 --> 00:01:29,560
However, when these
type 5 or type 3

29
00:01:29,560 --> 00:01:33,880
LSAs depending if they're
inter area or external routes,

30
00:01:33,880 --> 00:01:37,420
when these change
they disappear or they

31
00:01:37,420 --> 00:01:38,950
come into the routing table.

32
00:01:38,950 --> 00:01:41,620
That at least causes
everybody to have

33
00:01:41,620 --> 00:01:45,080
to update their database and
update their routing table.

34
00:01:45,080 --> 00:01:49,750
So let's begin by looking at
our topology diagram learning

35
00:01:49,750 --> 00:01:53,140
where our routes are coming from
that we're going to summarize.

36
00:01:53,140 --> 00:01:55,540
And then in this
video, we'll just

37
00:01:55,540 --> 00:01:57,460
be looking at what
actually takes

38
00:01:57,460 --> 00:02:00,760
place when those
routes flap so when

39
00:02:00,760 --> 00:02:04,580
they disappear and reappear
in the routing table.

40
00:02:04,580 --> 00:02:08,539
So again, let's start by
looking at that topology.

41
00:02:08,539 --> 00:02:12,340
So what we've done for this lab
is we've configured router 10

42
00:02:12,340 --> 00:02:17,890
to send in four routes that are
pretty much within what would

43
00:02:17,890 --> 00:02:21,730
be a /16 subnet.

44
00:02:21,730 --> 00:02:26,770
All of these routes
start with 100.100.

45
00:02:26,770 --> 00:02:29,620
And then they are different
in the third octet.

46
00:02:29,620 --> 00:02:32,360
We'll be taking a look at this,
of course, in just a moment.

47
00:02:32,360 --> 00:02:35,320
We've done the same
thing on router 9.

48
00:02:35,320 --> 00:02:38,080
It is also bringing
in four routes

49
00:02:38,080 --> 00:02:41,950
into in this case
EIGRP which, of course,

50
00:02:41,950 --> 00:02:47,650
router 2 is then being the ASBR
and redistributing those routes

51
00:02:47,650 --> 00:02:49,360
into OSPF.

52
00:02:49,360 --> 00:02:52,810
So those four routes, of course,
are coming in as external.

53
00:02:52,810 --> 00:02:57,490
These all begin with 200.200.

54
00:02:57,490 --> 00:03:02,110
Again, these could be
summarized easily with a /16.

55
00:03:02,110 --> 00:03:04,000
So let's jump in.

56
00:03:04,000 --> 00:03:05,920
And we're of course,
going to start off

57
00:03:05,920 --> 00:03:08,680
by looking at router 7,
which is a router very

58
00:03:08,680 --> 00:03:13,450
comfortably inside of our OSPF
area 0 environment meaning

59
00:03:13,450 --> 00:03:16,670
it is not an ABR or
an ASBR in any way.

60
00:03:16,670 --> 00:03:19,300
It's just an internal
router in area 0.

61
00:03:19,300 --> 00:03:21,130
And what, of course,
we'll be looking

62
00:03:21,130 --> 00:03:25,330
for here is what these
routes look like on router 7

63
00:03:25,330 --> 00:03:30,040
and what happens if those
routes go away and come back,

64
00:03:30,040 --> 00:03:33,640
what happens with our
database, and running Dijkstra.

65
00:03:33,640 --> 00:03:37,000
So let's jump into the
command line on router 7

66
00:03:37,000 --> 00:03:39,280
and get started by
looking at these routes

67
00:03:39,280 --> 00:03:44,320
before we start taking them
away and bringing them back.

68
00:03:44,320 --> 00:03:46,590
So let's begin by
looking at those routes

69
00:03:46,590 --> 00:03:48,300
that we were just discussing.

70
00:03:48,300 --> 00:03:50,850
We'll just do a
show IP route OSPF.

71
00:03:50,850 --> 00:03:57,060
And what we see is this block
of four 100.100 networks.

72
00:03:57,060 --> 00:04:00,480
These are the ones that
are coming through router 6

73
00:04:00,480 --> 00:04:02,220
from router 10.

74
00:04:02,220 --> 00:04:04,240
And then down here
at the bottom,

75
00:04:04,240 --> 00:04:09,120
we have the x channels in
the 200.200 range coming from

76
00:04:09,120 --> 00:04:12,690
router 9 through router
2 and then, of course,

77
00:04:12,690 --> 00:04:14,070
through router 1.

78
00:04:14,070 --> 00:04:17,610
Let's also take a
look at how many times

79
00:04:17,610 --> 00:04:21,839
the spoof algorithm has
executed in this area.

80
00:04:21,839 --> 00:04:24,780
As of right now, it is 21.

81
00:04:24,780 --> 00:04:28,320
Let's also turn on
our d bugs for OSPF.

82
00:04:28,320 --> 00:04:32,670
So now let's go up to router
10 and cause a failure in one

83
00:04:32,670 --> 00:04:35,190
of those 100.100 networks.

84
00:04:35,190 --> 00:04:37,140
We'll just bring down
interface loopback.

85
00:04:37,140 --> 00:04:37,950
110.

86
00:04:37,950 --> 00:04:39,330
We'll simply do a shot.

87
00:04:39,330 --> 00:04:42,870
And let's go back and observe
what happens on router 7.

88
00:04:42,870 --> 00:04:45,060
So the first thing
we see here is

89
00:04:45,060 --> 00:04:47,040
we definitely see the change.

90
00:04:47,040 --> 00:04:54,780
We see that we're processing the
fact that this 100.100.110.0/24

91
00:04:54,780 --> 00:04:58,540
is now being removed
from our routing table.

92
00:04:58,540 --> 00:05:01,130
We do a show IP
router OSPF again.

93
00:05:01,130 --> 00:05:03,950
We see that in fact
that network is gone.

94
00:05:03,950 --> 00:05:05,350
We, of course,
have the remainder

95
00:05:05,350 --> 00:05:06,820
of the other three networks.

96
00:05:06,820 --> 00:05:08,290
But that one is missing.

97
00:05:08,290 --> 00:05:13,720
And notice that OSPF
has executed 21 times.

98
00:05:13,720 --> 00:05:17,770
Now, keep in mind this is
the behavior we would expect.

99
00:05:17,770 --> 00:05:22,600
Remember, the point of
splitting OSPF up into areas

100
00:05:22,600 --> 00:05:26,800
is to as so we don't have
to run OSPF every time

101
00:05:26,800 --> 00:05:29,450
something changes
in a different area.

102
00:05:29,450 --> 00:05:34,270
So this is expected either
with or without summarization.

103
00:05:34,270 --> 00:05:37,240
What we're going to be
looking for from summarization

104
00:05:37,240 --> 00:05:41,780
is for router 7 not to have
to process an update at all.

105
00:05:41,780 --> 00:05:45,170
So just so we're clear
on that difference.

106
00:05:45,170 --> 00:05:46,990
We're still getting an update.

107
00:05:46,990 --> 00:05:49,570
We're still having to
process that update,

108
00:05:49,570 --> 00:05:53,410
modify our database, and
update our routing table.

109
00:05:53,410 --> 00:05:56,380
What we are not
doing is running OSPF

110
00:05:56,380 --> 00:05:58,760
because it is a different area.

111
00:05:58,760 --> 00:06:00,850
So once we get
into summarization,

112
00:06:00,850 --> 00:06:05,260
we'll see how that changes and
we don't have to do anything.

113
00:06:05,260 --> 00:06:07,130
That's our goal.

114
00:06:07,130 --> 00:06:09,100
So now let's go
back to router 10.

115
00:06:09,100 --> 00:06:11,510
We'll go ahead and bring
that interface back up.

116
00:06:11,510 --> 00:06:13,810
So this is basically
acting like a flap

117
00:06:13,810 --> 00:06:15,910
of the network on router 10.

118
00:06:15,910 --> 00:06:18,100
And if we go back
over to router 7,

119
00:06:18,100 --> 00:06:22,060
we see that once again we
had to process this update.

120
00:06:22,060 --> 00:06:23,980
Did we have to run OSPF?

121
00:06:23,980 --> 00:06:27,130
No, we're still at 21 times.

122
00:06:27,130 --> 00:06:30,700
However, we did have to process
this update that came in

123
00:06:30,700 --> 00:06:33,010
for our database
and, of course, that

124
00:06:33,010 --> 00:06:35,300
means we had to update
our routing table.

125
00:06:35,300 --> 00:06:41,200
Now we have that 100.100.110.0
back in our routing table as we

126
00:06:41,200 --> 00:06:41,740
should.

127
00:06:41,740 --> 00:06:44,868
We'll see that external routes
are processed the same way.

128
00:06:44,868 --> 00:06:46,660
Let's just go ahead
and demonstrate it just

129
00:06:46,660 --> 00:06:47,990
for completeness.

130
00:06:47,990 --> 00:06:49,600
Let's jump over to router 9.

131
00:06:49,600 --> 00:06:52,310
And we'll shut down one of
these loopbacks as well.

132
00:06:52,310 --> 00:06:56,650
So for example, interface
loopback 210, we'll do a shot.

133
00:06:56,650 --> 00:06:59,260
Once again, hop back
over to router 7.

134
00:06:59,260 --> 00:07:02,020
We see once again it
has to process this.

135
00:07:02,020 --> 00:07:03,910
It looks a little bit
different because it's

136
00:07:03,910 --> 00:07:07,570
an external route however,
it's the same basic process.

137
00:07:07,570 --> 00:07:09,820
If we look at our show
IP route for OSPF,

138
00:07:09,820 --> 00:07:11,470
we see that route is missing.

139
00:07:11,470 --> 00:07:15,520
And if we look at SPF we see
that it has still only executed

140
00:07:15,520 --> 00:07:18,520
21 times here in area 0.

141
00:07:18,520 --> 00:07:21,050
So although it looks a
little bit different,

142
00:07:21,050 --> 00:07:23,740
this is functioning
exactly the same way

143
00:07:23,740 --> 00:07:28,337
as the type 3 LSA
coming from area 200.

144
00:07:28,337 --> 00:07:29,920
Let's go ahead and
bring that back up.

145
00:07:29,920 --> 00:07:36,890

146
00:07:36,890 --> 00:07:38,300
We look at router 7 again.

147
00:07:38,300 --> 00:07:40,880
We see that it has again
learned that update.

148
00:07:40,880 --> 00:07:44,410
We still have not run
OSPF any additional times.

149
00:07:44,410 --> 00:07:50,240
And our route is back
in our routing table.

150
00:07:50,240 --> 00:07:55,730
In this video, we took a look
at the two major disadvantages

151
00:07:55,730 --> 00:07:58,970
to not running summarization.

152
00:07:58,970 --> 00:08:02,910
First, we have a large number
of routes in our routing table.

153
00:08:02,910 --> 00:08:05,300
Now, in this case,
it's only eight routes.

154
00:08:05,300 --> 00:08:08,340
We're bringing in 4 from one
router and 4 from another.

155
00:08:08,340 --> 00:08:11,010
So it's not a really
large number of routes,

156
00:08:11,010 --> 00:08:14,720
but it is 8 where it
could possibly be 2.

157
00:08:14,720 --> 00:08:17,870
And more importantly,
we took a look

158
00:08:17,870 --> 00:08:20,630
at how each router
in every area.

159
00:08:20,630 --> 00:08:23,510
We focused on router 7
but all the other routers

160
00:08:23,510 --> 00:08:24,660
would be the same.

161
00:08:24,660 --> 00:08:26,900
They have to process
these updates

162
00:08:26,900 --> 00:08:29,220
and modify their routing table.

163
00:08:29,220 --> 00:08:32,570
Do they have to run
Dijkstra or OSPF algorithm?

164
00:08:32,570 --> 00:08:36,950
No, that's taken care of
by the areas themselves.

165
00:08:36,950 --> 00:08:39,080
But it is still a
little bit of work

166
00:08:39,080 --> 00:08:40,799
that has to be done
on the routers.

167
00:08:40,799 --> 00:08:42,799
And when you're talking
about scaling up

168
00:08:42,799 --> 00:08:47,450
a protocol like OSPF,
this becomes critical.

169
00:08:47,450 --> 00:08:49,670
I hope this has been
informative for you.

170
00:08:49,670 --> 00:08:52,810
And I'd like to thank
you for viewing.

171
00:08:52,810 --> 00:08:56,000