1
00:00:00,000 --> 00:00:00,960

2
00:00:00,960 --> 00:00:04,320
[AUDIO LOGO]

3
00:00:04,320 --> 00:00:08,160

4
00:00:08,160 --> 00:00:10,980
Now, about that loop.

5
00:00:10,980 --> 00:00:13,440
We already saw
that we've created

6
00:00:13,440 --> 00:00:18,300
a routing loop because of our
redistribution between OSPF

7
00:00:18,300 --> 00:00:20,220
and IS-IS.

8
00:00:20,220 --> 00:00:22,750
We also talked about
the overall problem.

9
00:00:22,750 --> 00:00:25,410
Although, we will review
it in just a moment.

10
00:00:25,410 --> 00:00:30,180
The problem is, we have our edge
routers, in this case, router 1

11
00:00:30,180 --> 00:00:33,220
and 3, that are not following.

12
00:00:33,220 --> 00:00:35,820
Let's call them our
native protocols.

13
00:00:35,820 --> 00:00:38,340
In other words, we
have a route that

14
00:00:38,340 --> 00:00:43,470
should be in IS-IS that
router is 3 actually learning

15
00:00:43,470 --> 00:00:45,150
from OSPF.

16
00:00:45,150 --> 00:00:48,420
Now, there are many
ways to prevent this,

17
00:00:48,420 --> 00:00:51,270
and we're certainly not
going to go over all of them.

18
00:00:51,270 --> 00:00:54,630
Most of them involve some
form of route filtering.

19
00:00:54,630 --> 00:00:57,840
And we've discussed route
filtering quite a bit already.

20
00:00:57,840 --> 00:01:00,340
We can use any of those methods.

21
00:01:00,340 --> 00:01:04,230
However, the most common
way to deal with this

22
00:01:04,230 --> 00:01:08,680
is to do route filtering
based on route tags.

23
00:01:08,680 --> 00:01:11,790
Now, like I said, there are
many other ways to handle this.

24
00:01:11,790 --> 00:01:14,140
But that's the one we're
going to focus on here.

25
00:01:14,140 --> 00:01:18,360
So let's bring up our topology--
quick reminder of what it

26
00:01:18,360 --> 00:01:20,430
is we need to block and filter.

27
00:01:20,430 --> 00:01:22,650
And then we'll jump
into the command line,

28
00:01:22,650 --> 00:01:27,660
implement our filter, and
see how it fixes our loop.

29
00:01:27,660 --> 00:01:31,640
So as a reminder, we have router
8 sending a route to router 7.

30
00:01:31,640 --> 00:01:38,180
Router 7 is then sending this
to router 1 and to router 3.

31
00:01:38,180 --> 00:01:41,600
These show up with an
administrative distance of 115

32
00:01:41,600 --> 00:01:44,900
because these are from IS-IS.

33
00:01:44,900 --> 00:01:50,670
And then one of the two routers
is going to redistribute first.

34
00:01:50,670 --> 00:01:54,270
Truth be told, we don't really
know which router it will be.

35
00:01:54,270 --> 00:01:56,310
If they were to reload,
the other router

36
00:01:56,310 --> 00:01:57,610
would start doing it.

37
00:01:57,610 --> 00:01:59,790
So our solution
that we put in place

38
00:01:59,790 --> 00:02:02,310
is going to have to
be on both routers.

39
00:02:02,310 --> 00:02:04,980
In this particular
case, it's router 1.

40
00:02:04,980 --> 00:02:07,620
But that's only because we've
observed what the routes look

41
00:02:07,620 --> 00:02:09,539
like that we know that.

42
00:02:09,539 --> 00:02:12,150
Again, it could be
either router that

43
00:02:12,150 --> 00:02:15,600
learns the route from IS-IS
first and redistributes it

44
00:02:15,600 --> 00:02:17,580
into OSPF.

45
00:02:17,580 --> 00:02:23,070
So again, in our example, router
1 redistributes this into OSPF.

46
00:02:23,070 --> 00:02:28,570
And then that route gets
learned through OSPF on router

47
00:02:28,570 --> 00:02:34,050
3, which, of course, had
an admin distance of 110.

48
00:02:34,050 --> 00:02:38,580
So ultimately, router 3
replaces its IS-IS route

49
00:02:38,580 --> 00:02:43,170
and redistributes the
OSPF back into IS-IS,

50
00:02:43,170 --> 00:02:46,650
creating our loop and
our route feedback.

51
00:02:46,650 --> 00:02:52,630
What we need to do is stop
this feedback from occurring.

52
00:02:52,630 --> 00:02:55,260
The best way to do
that is to configure

53
00:02:55,260 --> 00:03:01,650
router 3 to discard this route
and not accept it from OSPF.

54
00:03:01,650 --> 00:03:05,875
The most common way to
do that is with tags.

55
00:03:05,875 --> 00:03:07,500
Now, the good news
is we've already set

56
00:03:07,500 --> 00:03:09,210
ourselves up to do this.

57
00:03:09,210 --> 00:03:11,670
When we configured
our redistribution

58
00:03:11,670 --> 00:03:18,060
of IS-IS and OSPF, we set this
up to already tag the routes

59
00:03:18,060 --> 00:03:19,800
as 115.

60
00:03:19,800 --> 00:03:23,310
So all we have to do is
configure our edge routers,

61
00:03:23,310 --> 00:03:25,980
router 3 to stop
the current loop,

62
00:03:25,980 --> 00:03:29,680
router 1 just in case
router wins next time.

63
00:03:29,680 --> 00:03:31,680
So again, we want
to do this on both.

64
00:03:31,680 --> 00:03:33,570
But we just need
to configure these

65
00:03:33,570 --> 00:03:39,630
to not allow routes tagged
with 115 to come in to OSPF.

66
00:03:39,630 --> 00:03:44,900
So let's jump into our command
line and configure that.

67
00:03:44,900 --> 00:03:46,870
So we'll start here on router 1.

68
00:03:46,870 --> 00:03:50,530
What we need to do is
create a route policy.

69
00:03:50,530 --> 00:03:55,060
Let's call it RP_OSPF_FILTER.

70
00:03:55,060 --> 00:04:00,790
What it needs to say is
if the tag equals 115

71
00:04:00,790 --> 00:04:03,280
because that's what we're
tagging our routes as,

72
00:04:03,280 --> 00:04:07,060
then we want to drop the route.

73
00:04:07,060 --> 00:04:10,450
Else, of course, we
want to pass the route.

74
00:04:10,450 --> 00:04:13,970
We'll end our if, and
we'll end our policy.

75
00:04:13,970 --> 00:04:16,850
Then all we have to do
is apply the filter.

76
00:04:16,850 --> 00:04:20,440
So we'll go into our
address-family under OSPF 1.

77
00:04:20,440 --> 00:04:23,480
We'll apply this as
a distribute-list.

78
00:04:23,480 --> 00:04:29,590
So simply say distribute-list
route-policy RP_OSPF_FILTER,

79
00:04:29,590 --> 00:04:32,050
and that will be in.

80
00:04:32,050 --> 00:04:37,300
We'll commit those changes
and exit configuration.

81
00:04:37,300 --> 00:04:39,690
Now, remember, the
problem isn't really here.

82
00:04:39,690 --> 00:04:43,860
This is just putting in the
solution on router 1 in case

83
00:04:43,860 --> 00:04:47,100
the problem were to flip-flop
in the event of maybe

84
00:04:47,100 --> 00:04:50,130
routers reloading or
something along those lines.

85
00:04:50,130 --> 00:04:53,130
What we really need to do
is put this on router 3.

86
00:04:53,130 --> 00:04:56,160
And as we do so, let's
investigate a little bit more

87
00:04:56,160 --> 00:05:00,300
what's going on router 3 before
we just throw the solution on.

88
00:05:00,300 --> 00:05:05,220
So over on router 3, let's
just start with a show ip route

89
00:05:05,220 --> 00:05:06,960
isis.

90
00:05:06,960 --> 00:05:11,520
Notice, we're not getting
any routes from IS-IS.

91
00:05:11,520 --> 00:05:15,030
This is because if I
say show ip route ospf,

92
00:05:15,030 --> 00:05:20,460
I'm literally learning about
the entire network from OSPF.

93
00:05:20,460 --> 00:05:23,520
Of course, router 3
being an edge router,

94
00:05:23,520 --> 00:05:27,000
it shouldn't be learning
everything from OSPF.

95
00:05:27,000 --> 00:05:29,200
This is entirely the problem.

96
00:05:29,200 --> 00:05:31,530
So now, let's go ahead
and put our filter in.

97
00:05:31,530 --> 00:05:33,540
Since we've already
seen the filter,

98
00:05:33,540 --> 00:05:36,400
we'll just go ahead
and copy and paste it.

99
00:05:36,400 --> 00:05:38,880
So again, we're just pasting
that exact same policy here

100
00:05:38,880 --> 00:05:40,150
on router 3.

101
00:05:40,150 --> 00:05:43,980
We'll give it just a moment
for the protocols to converge.

102
00:05:43,980 --> 00:05:47,160
We can check it by doing
a show ip route isis.

103
00:05:47,160 --> 00:05:49,690
And let's see if we
have any routes now.

104
00:05:49,690 --> 00:05:52,350
So we can see that we are now
learning some of the routes

105
00:05:52,350 --> 00:05:56,040
from IS-IS, which is good.

106
00:05:56,040 --> 00:05:58,320
Again, keep in mind,
we're only going

107
00:05:58,320 --> 00:06:03,750
to be learning the routes from
IS-IS that are tagged as 115.

108
00:06:03,750 --> 00:06:08,910
So we'll still be getting some
of the other routes from OSPF.

109
00:06:08,910 --> 00:06:11,340
So if we do a show
ip route ospf,

110
00:06:11,340 --> 00:06:14,250
we can see that we still have
a large number of routes here.

111
00:06:14,250 --> 00:06:16,360
Some of them even
being externals.

112
00:06:16,360 --> 00:06:18,870
So some of those are
coming from the EIGRP.

113
00:06:18,870 --> 00:06:22,410
And some of those are coming
from the RIP domain, in fact.

114
00:06:22,410 --> 00:06:25,990
But let's go ahead and
check our loop again now.

115
00:06:25,990 --> 00:06:28,350
So if we go over to
router 5, remember

116
00:06:28,350 --> 00:06:32,460
that this is the original
test that we saw of a loop on.

117
00:06:32,460 --> 00:06:35,940
What we see here is now that
we've implemented our solution,

118
00:06:35,940 --> 00:06:39,780
not only do we not have
a loop, but we're also

119
00:06:39,780 --> 00:06:43,980
properly multipathing
through both edge routers.

120
00:06:43,980 --> 00:06:48,330
The first hop, of course, is
going between routers 2 and 4.

121
00:06:48,330 --> 00:06:53,340
And then number 2, we can
see we're using both edges.

122
00:06:53,340 --> 00:06:58,410
It's going through router 3 and
router 1 on that second hop.

123
00:06:58,410 --> 00:07:00,210
Now, of course, the
third half, ultimately,

124
00:07:00,210 --> 00:07:01,560
just goes to router 7.

125
00:07:01,560 --> 00:07:03,840
And then, of course,
the fourth to router 8

126
00:07:03,840 --> 00:07:06,630
because there's no multipathing
on that side of the network.

127
00:07:06,630 --> 00:07:11,870
But our OSPF environment is
now multipathing correctly.

128
00:07:11,870 --> 00:07:16,180
So in this video, we
explored the most popular way

129
00:07:16,180 --> 00:07:18,860
to solve routing loop issues.

130
00:07:18,860 --> 00:07:21,190
Now, as I mentioned
during the introduction,

131
00:07:21,190 --> 00:07:23,480
there are other ways to filter.

132
00:07:23,480 --> 00:07:25,450
There are some other
ways you can do things,

133
00:07:25,450 --> 00:07:28,510
like using split
administrative distance where

134
00:07:28,510 --> 00:07:32,090
your internal routes have a
lower administrative distance.

135
00:07:32,090 --> 00:07:35,820
And your external routes have a
higher administrative distance.

136
00:07:35,820 --> 00:07:39,580
EIGRP, for example,
does that by default.

137
00:07:39,580 --> 00:07:41,980
OSPF is also
capable of doing it.

138
00:07:41,980 --> 00:07:44,240
Although, it's not
enabled by default.

139
00:07:44,240 --> 00:07:46,540
The only problem with
solutions like that

140
00:07:46,540 --> 00:07:50,650
is they really only work if
they're on the network edge.

141
00:07:50,650 --> 00:07:55,510
They cannot be a transit
domain for our routing and have

142
00:07:55,510 --> 00:07:57,530
the split admin distance work.

143
00:07:57,530 --> 00:08:00,490
One of the advantages
to doing tags like this

144
00:08:00,490 --> 00:08:03,850
is with a proper
design, this will always

145
00:08:03,850 --> 00:08:05,680
solve your routing loops.

146
00:08:05,680 --> 00:08:08,890
And as you can see, when
we solve those loops,

147
00:08:08,890 --> 00:08:11,350
we can also pick
up multipathing.

148
00:08:11,350 --> 00:08:15,640
So that concludes everything we
need to discuss in this skill.

149
00:08:15,640 --> 00:08:18,520
That leaves us
with just a review

150
00:08:18,520 --> 00:08:22,220
in the form of a quiz, which
we're going to do next.

151
00:08:22,220 --> 00:08:23,300
Until then.

152
00:08:23,300 --> 00:08:25,450
I hope this has been
informative for you,

153
00:08:25,450 --> 00:08:28,500
and I'd like to thank
you for viewing.

154
00:08:28,500 --> 00:08:32,000