WEBVTT

0:00:08.660000 --> 0:00:17.300000
 All right. Next we're going to look
 at IPv6 over an IPv4 TCP session.

0:00:17.300000 --> 0:00:21.920000
 So again, we're just going to jump
 right to the command line and we're

0:00:21.920000 --> 0:00:23.800000
 going to start with router 3.

0:00:23.800000 --> 0:00:26.940000
 Now we're going to be building this
 in the same areas we were with the

0:00:26.940000 --> 0:00:29.060000
 previous lessons.

0:00:29.060000 --> 0:00:32.700000
 So again, if you go to the diagram,
 in fact let me bring that up here

0:00:32.700000 --> 0:00:36.140000
 real quick for us since we haven't looked
 at it here for a little while.

0:00:36.140000 --> 0:00:41.680000
 So we're going to be building this again
 up between router 3 and switch

0:00:41.680000 --> 0:00:46.280000
 3. And possibly if we have some stuff
 we need to look at between switch

0:00:46.280000 --> 0:00:50.400000
 3 and router 6. But probably not.

0:00:50.400000 --> 0:00:54.700000
 We'll probably see everything we need
 to just on one BGP session here.

0:00:54.700000 --> 0:00:59.220000
 I don't think we have to do a whole
 lot with it to take a look at this.

0:00:59.220000 --> 0:01:02.200000
 And we'll see some of this a little
 bit more later to this particular

0:01:02.200000 --> 0:01:09.040000
 topic. So as far as that goes, I don't
 think we need to actually do anything

0:01:09.040000 --> 0:01:15.680000
 to see what we need to see rather
 than just router 3 and switch 3.

0:01:15.680000 --> 0:01:16.780000
 So let's take a look.

0:01:16.780000 --> 0:01:22.060000
 So on router 3 we're going to, and again,
 between the lessons here I am

0:01:22.060000 --> 0:01:24.540000
 taking out the BGP we had in.

0:01:24.540000 --> 0:01:27.360000
 So we're going to say
 router BGP 100.

0:01:27.360000 --> 0:01:34.300000
 And in this case, just sort of to show
 that we're not sending, or that

0:01:34.300000 --> 0:01:40.060000
 we don't have to send address family
 IPV4 at all, we could turn it off.

0:01:40.060000 --> 0:01:44.900000
 We could say no BGP default IPV4.

0:01:44.900000 --> 0:01:50.960000
 And then address family
 IPV6 unicast.

0:01:50.960000 --> 0:01:56.420000
 Neighbor. And I know I don't have the
 addresses shown on our IPV6 diagram.

0:01:56.420000 --> 0:02:05.140000
 But if I want to appear between physicals
 up there, it would be 50.1.39

0:02:05.140000 --> 0:02:11.900000
.9. Remote AS 300.

0:02:11.900000 --> 0:02:15.500000
 And let's see what we get.

0:02:15.500000 --> 0:02:21.520000
 So now what we're going to
 have is turning off IPV4.

0:02:21.520000 --> 0:02:25.720000
 Again, it moves the neighbor statement
 up to where it should be and puts

0:02:25.720000 --> 0:02:28.560000
 the activate under IPV6.

0:02:28.560000 --> 0:02:35.060000
 And again, if we said do show TCP, brief
 all, it's now listening for an

0:02:35.060000 --> 0:02:38.040000
 IPV4 session between
 these devices.

0:02:38.040000 --> 0:02:42.060000
 So let's go to switch three.

0:02:42.060000 --> 0:02:47.400000
 So router BGP 300.

0:02:47.400000 --> 0:02:51.640000
 No BGP default IPV4.

0:02:51.640000 --> 0:02:57.200000
 Address family IPV6 unicast.

0:02:57.200000 --> 0:03:05.260000
 Neighbor 50.1.39.3.

0:03:05.260000 --> 0:03:11.760000
 Remote AS 100. Neighbor comes up.

0:03:11.760000 --> 0:03:15.260000
 And as before, we need
 to give it a network.

0:03:15.260000 --> 0:03:20.020000
 So let's have it inject
 his loop back.

0:03:20.020000 --> 0:03:33.720000
 There we go. Do show
 BGP IPV6 unicast.

0:03:33.720000 --> 0:03:39.360000
 And just like before, he's injecting
 his network into BGP.

0:03:39.360000 --> 0:03:44.060000
 No big deal. Where things get a little
 more interesting with this setup

0:03:44.060000 --> 0:03:53.800000
 is as we go to router three and
 say do show BGP IPV6 unicast.

0:03:53.800000 --> 0:03:59.280000
 The really fun part
 is the next hop.

0:03:59.280000 --> 0:04:05.800000
 This is referred to as
 a mapped IPV4 address.

0:04:05.800000 --> 0:04:10.800000
 Now there's something called an IPV4
 compatibility address as well, which

0:04:10.800000 --> 0:04:14.840000
 is the colon colon
 without the F's.

0:04:14.840000 --> 0:04:19.380000
 So it would just be colon colon
 and then the IPV4 address.

0:04:19.380000 --> 0:04:21.700000
 Not as popular anymore.

0:04:21.700000 --> 0:04:27.800000
 There can be some confusion issues as far
 as whether it's an IPV4 compatibility

0:04:27.800000 --> 0:04:34.300000
 address or if it's just an oddly
 formatted IPV6 address.

0:04:34.300000 --> 0:04:38.140000
 So these are becoming
 a little more popular.

0:04:38.140000 --> 0:04:43.980000
 This is the mapped IPV4 address where
 it puts simply the colon colon,

0:04:43.980000 --> 0:04:48.380000
 FFFF colon 50.1.39.9.

0:04:48.380000 --> 0:04:58.500000
 Now the problem of course is do show
 BGP IPV6 unicast for that prefix.

0:04:58.500000 --> 0:05:01.020000
 That was interesting.

0:05:01.020000 --> 0:05:07.400000
 Where do you know what
 this is going to say?

0:05:07.400000 --> 0:05:12.700000
 Inaccessible. He's like, I
 don't know what that is.

0:05:12.700000 --> 0:05:14.420000
 I don't know what to do with that.


0:05:14.420000 --> 0:05:18.120000
 I can't use that.

0:05:18.120000 --> 0:05:21.400000
 Now a lot of people might look at this
 and go, okay, we'll go to the other

0:05:21.400000 --> 0:05:23.480000
 router, do next hop self.

0:05:23.480000 --> 0:05:25.680000
 Well, this is EBGP.

0:05:25.680000 --> 0:05:30.220000
 He's already sort of
 doing next hop self.

0:05:30.220000 --> 0:05:34.760000
 So here's the thing.

0:05:34.760000 --> 0:05:39.800000
 That addressing and again, yeah, if some
 people look at it and go, that's

0:05:39.800000 --> 0:05:41.700000
 really stupid. Why
 is it doing that?

0:05:41.700000 --> 0:05:42.660000
 That doesn't make any sense.

0:05:42.660000 --> 0:05:45.040000
 He can't use that.

0:05:45.040000 --> 0:05:49.140000
 Agreed. He can't.

0:05:49.140000 --> 0:05:52.300000
 Not yet. Not yet.

0:05:52.300000 --> 0:05:59.660000
 But when we get to MPLS,
 specifically 6PE.

0:05:59.660000 --> 0:06:05.240000
 So when we get to 6PE, we're going
 to see those exact routes just like

0:06:05.240000 --> 0:06:07.760000
 that, just like we see right now.

0:06:07.760000 --> 0:06:13.740000
 But instead of them not being selected
 as best and saying inaccessible,

0:06:13.740000 --> 0:06:16.420000
 we'll be able to use them.

0:06:16.420000 --> 0:06:22.080000
 And the reason is, is because we'll be using
 MPLS to forward to the destination

0:06:22.080000 --> 0:06:28.700000
 rather than trying to use actual
 IP or IPV6 destination addresses.

0:06:28.700000 --> 0:06:34.140000
 So once MPLS is involved, we won't
 have to do anything special.

0:06:34.140000 --> 0:06:37.120000
 The router will already
 be using these.

0:06:37.120000 --> 0:06:42.060000
 So I'm just trying to point out that
 in this particular case, this is

0:06:42.060000 --> 0:06:43.740000
 not going to work for us.

0:06:43.740000 --> 0:06:46.060000
 This next hop is no good.

0:06:46.060000 --> 0:06:49.360000
 And going to the other side and saying
 next hop self, that's not going

0:06:49.360000 --> 0:06:52.660000
 to help us either, because that's
 what he's already doing.

0:06:52.660000 --> 0:06:56.480000
 So in this case, this isn't
 buying us anything.

0:06:56.480000 --> 0:06:59.020000
 It's not helping.

0:06:59.020000 --> 0:07:02.100000
 But I just want to point out, and we'll
 see it again later, but when we

0:07:02.100000 --> 0:07:07.500000
 get to MPLS, that next hop address is
 actually going to be perfectly fine.

0:07:07.500000 --> 0:07:09.200000
 There's nothing wrong with that.

0:07:09.200000 --> 0:07:15.360000
 So without MPLS though, how do we
 get this to work for our benefit?

0:07:15.360000 --> 0:07:16.780000
 And so it'll work right now.

0:07:16.780000 --> 0:07:20.000000
 It just means we're going to have
 to manually set the next hop.

0:07:20.000000 --> 0:07:22.760000
 So let's go to switch three.

0:07:22.760000 --> 0:07:34.580000
 And we will say, route map, Rm2R3.


0:07:34.580000 --> 0:07:51.960000
 4v6. And I will say set IPv6 next hop
 to 2001 DB8 colon 100 colon 39 colon

0:07:51.960000 --> 0:07:58.320000
 colon 9, which is our address.

0:07:58.320000 --> 0:08:05.760000
 Router BGP 300. Add just
 family IPv6 unicast.

0:08:05.760000 --> 0:08:13.500000
 Neighbor 50 dot 1 dot 39 dot 3.

0:08:13.500000 --> 0:08:31.160000
 Route map. Out. Do clear BGP
 IPv6 unicast star out.

0:08:31.160000 --> 0:08:37.940000
 And we go to router
 3 and check it out.

0:08:37.940000 --> 0:08:41.800000
 See, much better.

0:08:41.800000 --> 0:08:44.220000
 All happy. So I set
 it to that address.

0:08:44.220000 --> 0:08:49.240000
 And now this looks exactly
 like it did.

0:08:49.240000 --> 0:08:54.200000
 Back we'll be redoing IPv6
 over an IPv6 session.

0:08:54.200000 --> 0:09:08.820000
 The difference is, if I say do show TCP
 brief all, that it's an IPv4 session

0:09:08.820000 --> 0:09:19.180000
 that's established.

0:09:19.180000 --> 0:09:23.100000
 Yet we're learning our
 BGP route just fine.

0:09:23.100000 --> 0:09:26.320000
 And we should be able to ping it
 since we're directly connected.

0:09:26.320000 --> 0:09:31.380000
 We shouldn't have next hop reach ability
 issues or return traffic issues.

0:09:31.380000 --> 0:09:32.800000
 It should just work.

0:09:32.800000 --> 0:09:35.480000
 And not with a mask on it though.

0:09:35.480000 --> 0:09:43.620000
 Here we go. So one more time.

0:09:43.620000 --> 0:09:48.660000
 IPv6 over IPv4. Not
 really a big deal.

0:09:48.660000 --> 0:09:58.740000
 Let me show you the running
 config though on switch 3.

0:09:58.740000 --> 0:10:05.040000
 Probably the trickiest thing to remember
 is that you're going under IPv6

0:10:05.040000 --> 0:10:11.140000
 address family and then your neighbor you're
 putting in with an IPv4 address.

0:10:11.140000 --> 0:10:15.740000
 This is what determines what the session
 is going to be established over.

0:10:15.740000 --> 0:10:19.900000
 But then of course we're still going
 to inject IPv6 addresses and inject

0:10:19.900000 --> 0:10:22.900000
 IPv6 addresses over the sessions.

0:10:22.900000 --> 0:10:28.140000
 The big gotcha, the big watch it just
 like in our previous lesson of IPv6

0:10:28.140000 --> 0:10:34.380000
 over IPv6. The big gotcha is again
 watching for that next hop.

0:10:34.380000 --> 0:10:40.200000
 If it's not reachable, then you have
 to do a route map and just set the

0:10:40.200000 --> 0:10:40.940000
 next hop manually.