WEBVTT

00:00:00.000 --> 00:00:06.240
Welcome back to vSphere 8 in the real world, and now we're going to talk about setting up your networking.

00:00:06.240 --> 00:00:11.940
So, one of the things you're going to want to do is you're going to need to add an adapter, a physical adapter, to your physical switches.

00:00:11.940 --> 00:00:20.500
Now, first you need to create a physical switch, and once you've created a physical switch, you can go in there and you can enable the adapter that you want to assign to that switch.

00:00:20.500 --> 00:00:27.000
You can either make it an unclaimed adapter, you can make it an active adapter, a standby adapter, or an unused.

00:00:27.000 --> 00:00:39.500
Unclaimed means that the adapter isn't assigned to any virtual switch, and unused means it's assigned to the switch, but currently it's not set to be an active or a standby, and then all you do is just move it up.

00:00:39.500 --> 00:00:47.400
If it's assigned to a switch but unused, then another switch can't use it. If it's unclaimed, it can be assigned to a switch.

00:00:47.400 --> 00:00:52.299
A physical NIC on a host can only belong to one virtual switch.

00:00:52.299 --> 00:01:06.900
Now, that virtual switch might have multiple port groups, so you can have multiple different policies for that NIC based on what port group it's assigned to, based on the traffic coming from that port group, but a NIC can only belong to one virtual switch.

00:01:06.900 --> 00:01:17.300
So, if it's unclaimed, it can be assigned to another virtual switch. If it's claimed and unused, it's still claimed by that switch, and so another switch can't claim it.

00:01:17.300 --> 00:01:24.199
For your adapters, as it's showing here, you're going to go to the host, you're going to go to Networking, Virtual Switches, select the switch and say Manage Adapters.

00:01:24.199 --> 00:01:31.099
So, there is when you'd go and you would claim that adapter. So, in order to do this, you'd have to already have a virtual switch created.

00:01:31.099 --> 00:01:38.000
So, I know it says it's Step 1, but it's Step 1 after you've created your switch, and the next step we're going to talk about is creating the switch.

00:01:38.000 --> 00:01:44.199
So, you can go to your adapters, you can add an adapter, and then you can set your policy for that adapter.

00:01:44.199 --> 00:01:54.800
All right. So, there you're seeing, select your target device. So, we're going to say, are we going to create a new virtual switch, or are we going to use an existing virtual switch?

00:01:54.800 --> 00:02:00.300
If we want to just manage the adapter and assign it to a switch, we can then go and say we want to create a virtual switch.

00:02:00.300 --> 00:02:06.500
It's like when we create a port group. It can say, do you want to create a new port group for an existing switch, or do you want to create a new switch, right?

00:02:06.500 --> 00:02:19.800
Or you can also create that switch in advance. I think most likely you'd create your virtual switch first, get any configuration you want on that, and assign your adapters to that switch with all the configuration for the switch on there.

00:02:19.800 --> 00:02:31.100
So, here we're talking about creating a switch. All right. So, here we're going to go into our host, because if we're going to the host and creating a switch, we're creating a standard virtual switch.

00:02:31.100 --> 00:02:36.800
We'll talk about distributed virtual switches in a little bit where we'd go into vSphere and create one for the data center.

00:02:36.800 --> 00:02:46.500
But right now, we're creating standard virtual switches, which means we're going to the host and then going to their network configuration, and we're creating our virtual switch.

00:02:46.500 --> 00:02:54.300
So, we're going to say we're going to configure our networking virtual switches, and then we can add and we can create that switch.

00:02:54.300 --> 00:03:05.000
We can go in, we can give it a name, we can assign it a VLAN tag. There's other things we can do as well for a virtual switch, and we'll talk about, like, policies and things like that coming up.

00:03:05.000 --> 00:03:18.100
And then we could, you know, finish that. Now, once you have created your virtual switch and probably assigned a physical adapter to it, the next thing you'll probably want to do is create a port group.

00:03:18.100 --> 00:03:23.000
And there's a lot of settings, and usually we create port groups if we want to change all these settings.

00:03:23.000 --> 00:03:29.000
They can inherit settings from their virtual switch, or you can override certain ones.

00:03:29.000 --> 00:03:36.699
So, let's say, for example, for security, we have a standard security policy that we want all of our port groups to use without exception.

00:03:36.699 --> 00:03:45.300
We do not allow forged transmit. We do not allow promiscuous mode. We do not allow MAC address spoofing.

00:03:45.300 --> 00:03:50.300
We don't allow any of those things, any on our network at all, ever.

00:03:50.300 --> 00:03:58.399
So, then I can create that policy on the virtual switch, and all my port groups are going to inherit that, and I just don't override that policy on the port group.

00:03:58.399 --> 00:04:05.500
But maybe I am going to change the adapter settings. I'll have some adapters be active, some be standby. They'll be different.

00:04:05.500 --> 00:04:10.399
Each port group might have a different VLAN tag on each of them, so that would be different.

00:04:10.600 --> 00:04:16.000
So, I can change certain things, and I can leave certain things the same, but the things are going to be the same.

00:04:16.000 --> 00:04:20.200
I can set on a switch level and not have to even worry about them on the port group level.

00:04:20.200 --> 00:04:25.200
It gets inherited down through all of them. All right.

00:04:25.200 --> 00:04:29.899
So, here we're seeing port groups. Now, this is for a VM kernel port.

00:04:29.899 --> 00:04:36.799
So, here I can give the VM kernel port a name, and maybe I'm creating this VM kernel port for my vMotion network.

00:04:36.799 --> 00:04:43.600
I'm going to have two adapters on my host that I'm dedicating just to vMotion traffic, and I don't want any other traffic on those.

00:04:43.600 --> 00:04:51.600
So, I can create a virtual switch for this if I wanted to, and then I can go in and create a VM kernel port on that switch that uses those adapters.

00:04:51.600 --> 00:04:58.299
If I am okay with it going over the same network as my virtual machines, not usually considered a best practice,

00:04:58.299 --> 00:05:04.500
but if I am okay with that, I can use the same virtual switch that has my virtual machines on it,

00:05:04.500 --> 00:05:09.399
and then I can create a VM kernel port on that switch, give it a name.

00:05:09.399 --> 00:05:13.299
Maybe in this situation, I'd call it the vMotion, you know, kernel port.

00:05:13.299 --> 00:05:18.899
I can give it a VLAN tag. My vMotion network is going to be VLAN 150.

00:05:18.899 --> 00:05:25.600
If I'm going to set up IPv4, which I'm sure I will set up IPv4 on there, if I'm going to use a different MTU size.

00:05:25.600 --> 00:05:33.299
So, if I change the MTU here, if I don't use the standard default 1,500, I want to go to 9,000 and use jumbo frames instead,

00:05:33.299 --> 00:05:38.100
I need to make sure I've enabled jumbo frames on my physical switch as well.

00:05:38.100 --> 00:05:46.799
My NIC needs to be, you know, able to accept jumbo frames of 9,000, and my switch port needs to be configured for that as well.

00:05:46.799 --> 00:05:53.899
Nowadays, it's pretty common where people go in there and they'll set the MTU on all their switch ports to like 9,100 or something like that,

00:05:53.899 --> 00:05:58.500
and that way they can enable jumbo frames when they need to, when it's appropriate for them.

00:05:58.500 --> 00:06:03.399
Then I'm going to have to say what services for this VM kernel port I want to use.

00:06:03.399 --> 00:06:07.899
If I don't check any of these boxes, then this VM kernel port is just a management VM kernel port.

00:06:07.899 --> 00:06:16.000
I can use it to manage the host. I can use it to connect to vSphere, or I can connect to that IP address and manage the host directly.

00:06:16.000 --> 00:06:23.200
If I want to use it for something else, like say, for example, for vMotion, then I'm going to check that vMotion box, and now I can use it for vMotion.

00:06:23.200 --> 00:06:32.200
If I want to use it for fault-tolerant VMs, right, for the machines that manage fault-tolerant VMs across the host, then I could check fault-tolerant logging, things like that, okay?

00:06:32.200 --> 00:06:39.200
So, you'll have to go in there and enable the services. vSAN, obviously, another one that's really common that probably gets checked quite a bit.

00:06:39.200 --> 00:06:44.399
So, it depends on what services you're using for that kernel port, you can configure it here.

00:06:44.399 --> 00:06:51.700
So, this is how you would also create a VM kernel port just for vMotion, and you might create another one just for vSAN.

00:06:51.700 --> 00:06:59.200
And if you have iSCSI, then you would create a VM kernel port for iSCSI and assign that VM kernel port to your iSCSI software initiator, all right?

00:06:59.200 --> 00:07:09.200
So, we'll cover that when we get, you know, further into this lessons, but you'll see coming up, you can create different types of VM kernel ports based upon different services you want to use those kernel ports for.

00:07:12.200 --> 00:07:17.200
So, adding a VM kernel port, so after you've, you know, you have a switch, you have the switch configured you want.

00:07:17.200 --> 00:07:29.700
So, if we're going to, in the example I just mentioned, if we want to have an iSCSI network with iSCSI adapters that are dedicated just for iSCSI, maybe you have a SmartNIC that has iSCSI offloading.

00:07:29.700 --> 00:07:43.700
And so, for those SmartNICs that have the iSCSI offloading, you want those dedicated to your iSCSI network, create a virtual switch, assign those two adapters to that virtual switch, create a VM kernel port on that virtual switch, okay?

00:07:44.200 --> 00:07:51.200
So, in here you see all the steps, you know, there's add networking, choose an adapter, select the switch, add your VLAN tag, set your IP address.

00:07:51.200 --> 00:07:58.200
If you're going to change the MTU, change all that stuff, right, choose what service you want, and then next and finish and review.

00:07:59.700 --> 00:08:01.700
So, all the things we saw in that last slide.

00:08:03.700 --> 00:08:10.200
So, here's going just to a quick demo of a, just like a standard virtual switch on an ESXi host.

00:08:10.700 --> 00:08:24.700
All right, here we are in our vSphere environment, and I'm going to go ahead and go to my Houston data center, and I have my prod cluster, and I have a host in my prod cluster, and one place that we can go to look at our networking is we can just go to networks.

00:08:24.700 --> 00:08:38.700
And from here, I can see I have a port group called VM networks, just on a standard, you know, network type, but what I want to do, here's just showing the vCenter server is on there.

00:08:39.200 --> 00:08:47.200
What I want to do is I want to go to configure, and under networking, I have my virtual switches.

00:08:47.200 --> 00:08:55.700
And here I can see I do have virtual switch zero, which has my management VLAN with a VM kernel port.

00:08:56.700 --> 00:08:58.200
I got my VM network.

00:08:58.200 --> 00:09:06.200
I've got a vMotion network with another VM kernel port, and I've got another switch, which I have, I call my storage switch.

00:09:09.200 --> 00:09:18.200
And I have an iSCSI VM kernel port on this storage switch, right?

00:09:18.200 --> 00:09:23.200
So, for this demo, let's just go ahead, and we're going to say we want to add networking.

00:09:25.200 --> 00:09:36.200
And I'm just going to create a virtual machine port group for a standard switch, and I can say I want to create a new standard switch, and I can set the MTU size.

00:09:36.700 --> 00:09:41.700
If I wanted to use jumbo frames, I would change this to 9,000, but I'll go ahead and leave it on the standard 1,500.

00:09:43.200 --> 00:09:45.200
Do I want to use any unclaimed adapters right now?

00:09:45.200 --> 00:09:46.200
Here's the unclaimed.

00:09:46.200 --> 00:09:48.700
Are there any active ones set on here?

00:09:48.700 --> 00:09:49.700
Nothing like that.

00:09:49.700 --> 00:09:50.700
All right.

00:09:50.700 --> 00:09:53.700
I can go ahead and click on this adapter and say move it down.

00:09:53.700 --> 00:09:55.700
That would be unused.

00:09:55.700 --> 00:09:58.200
Or I can leave it under unclaimed.

00:09:58.700 --> 00:10:06.700
The reason why I might create a switch and not assign an adapter to it, that would create what's called an internal-only virtual network.

00:10:06.700 --> 00:10:23.700
And an internal-only virtual network, what that means is VMs on this host specifically that are connected to this switch or to port groups on this switch can all talk to each other, but they can't talk outside of that network, only to each other.

00:10:23.700 --> 00:10:24.700
All right?

00:10:24.700 --> 00:10:28.200
So, I might leave it unclaimed and say next.

00:10:29.200 --> 00:10:31.200
I'm going to say no active adapters, okay.

00:10:32.700 --> 00:10:35.200
Do I want to give this a label?

00:10:35.200 --> 00:10:45.200
Maybe I'll go in there and just call this Demo 1.

00:10:45.200 --> 00:10:47.200
Maybe I'll put a VLAN tag on here if I want to.

00:10:47.200 --> 00:10:50.200
Maybe I'll put this on VLAN 200.

00:10:52.200 --> 00:10:53.200
Say next.

00:10:53.200 --> 00:11:00.200
Now, if I'm doing an internal-only virtual switch, you know, I'm using that going forward that way, I really don't need a VLAN tag.

00:11:00.200 --> 00:11:02.700
I can leave a no VLAN tag on there if I wanted to.

00:11:02.700 --> 00:11:03.700
I can say finish.

00:11:09.700 --> 00:11:10.700
And let's see.

00:11:13.700 --> 00:11:15.700
So, update network configuration.

00:11:15.700 --> 00:11:16.700
That says that's completed.

00:11:17.200 --> 00:11:18.200
Oh, there.

00:11:18.200 --> 00:11:19.200
Down there.

00:11:19.200 --> 00:11:20.200
I didn't scroll far enough.

00:11:20.200 --> 00:11:23.200
So, here is that new switch 1 that I created.

00:11:23.200 --> 00:11:26.200
And notice there's no adapters on this.

00:11:26.200 --> 00:11:30.200
So, if I wanted to, I could go in and say I want to manage physical adapters.

00:11:30.200 --> 00:11:33.200
I could choose that unclaimed NIC that's on there.

00:11:35.200 --> 00:11:38.200
I can see the status on there, speed, and all that kind of good stuff.

00:11:38.200 --> 00:11:39.200
All right?

00:11:39.200 --> 00:11:41.200
I'm going to go ahead and say choose that adapter.

00:11:41.700 --> 00:11:43.700
I'm going to go ahead and say choose that adapter.

00:11:43.700 --> 00:11:47.700
I'm going to move it down to my active adapters.

00:11:47.700 --> 00:11:49.700
I can say okay to that.

00:11:50.700 --> 00:11:53.700
And now I do have an active adapter on there.

00:11:55.700 --> 00:12:01.700
So, if I go down to my physical adapters here, I can see all the different adapters I have configured on this guy.

00:12:01.700 --> 00:12:09.700
Now, I added some more adapters real quick behind the scenes, a little virtualization magic, to this virtual host.

00:12:10.200 --> 00:12:11.200
Right?

00:12:11.200 --> 00:12:16.200
So, we had adapters 0 through 3, and I've added 4, 5, 6, and 7.

00:12:16.200 --> 00:12:20.200
So, now I can go back to my virtual switch, this guy here.

00:12:20.200 --> 00:12:23.200
I can say I want to manage physical adapters.

00:12:23.200 --> 00:12:28.200
Notice I have 4, 5, 6, and 7 unclaimed, and 2 is the active one that I chose during my setup.

00:12:28.200 --> 00:12:33.200
I'm going to go ahead and move that up to unclaimed.

00:12:33.200 --> 00:12:36.700
I'm going to move 4 down to standby.

00:12:36.700 --> 00:12:37.700
5.

00:12:42.700 --> 00:12:43.700
I had too many things selected.

00:12:43.700 --> 00:12:44.700
My apologies.

00:12:44.700 --> 00:12:46.700
Let me fix this real quick and move that back up.

00:12:46.700 --> 00:12:47.700
All right.

00:12:47.700 --> 00:13:02.700
Let me move 4, get rid of 2, move down, standby, and 5 down to 6 and 7, not 4, just 6 and 7.

00:13:03.200 --> 00:13:04.200
Move those down.

00:13:04.200 --> 00:13:12.200
So, now I have 6 and 7 active, and I have 4 and 5 are my standby for that particular switch.

00:13:15.200 --> 00:13:18.200
And notice that you see all the adapters here.

00:13:18.200 --> 00:13:25.200
So, now what I want to do is I want to create a port group on this switch.

00:13:25.200 --> 00:13:31.200
So, I'm going to say add networking, and I'm going to create a port group, and I'm going

00:13:31.200 --> 00:13:38.200
to create a port group on that new switch I just created, and I'm going to call this

00:13:38.200 --> 00:13:47.200
port group test port group 1.

00:13:47.200 --> 00:13:48.700
Maybe I'll add a VLAN tag to it.

00:13:48.700 --> 00:13:51.700
I'll add VLAN tag.

00:13:51.700 --> 00:13:57.200
If I can get this selected here, 210.

00:13:57.200 --> 00:13:59.700
Next, and I'm going to complete that.

00:13:59.700 --> 00:14:02.700
And I'm going to create another port group.

00:14:08.200 --> 00:14:16.200
Again, select an existing switch, test port group instead of 01.

00:14:16.200 --> 00:14:24.700
I want it to be 02, and I'll make it be 220 for that VLAN tag.

00:14:24.700 --> 00:14:26.200
Next, and finish.

00:14:26.200 --> 00:14:33.200
So, what I want to show you is between these two port groups, I have these four NICs selected,

00:14:33.200 --> 00:14:37.200
and I said 6 and 7 were my active, 4 and 5 are my standby.

00:14:37.200 --> 00:14:48.200
If I look at demo 1, view its settings, and I look at my failover, I can see my active

00:14:48.200 --> 00:14:55.700
are 6 and 7, and my standbys are 4 and 5, which is what my switch is configured for,

00:14:55.700 --> 00:15:01.700
and so my demo port group inherited those settings, right?

00:15:01.700 --> 00:15:08.200
Here, I want to go in, and I want to edit settings, and I'm going to go to traffic shaping,

00:15:08.200 --> 00:15:13.200
or no, sorry, teaming and failover, and I see that I have the same thing.

00:15:13.200 --> 00:15:19.200
My standbys are 4 and 5, and my actives are 6 and 7.

00:15:19.200 --> 00:15:24.700
Now, what I can do is say I want to override the failover order.

00:15:24.700 --> 00:15:30.200
So, this is the failover order it inherited from my switch, but for this port group,

00:15:30.200 --> 00:15:41.700
I actually want to move 6 and 7 down, and I want to move 4 and 5 to be active.

00:15:41.700 --> 00:15:43.200
So, they're the exact opposite.

00:15:43.200 --> 00:15:48.200
So, what happens is for one port group, the traffic's going over 6 and 7, and for the

00:15:48.200 --> 00:15:52.700
other port group, traffic's going over 4 and 5, and if everything's healthy and everything's

00:15:52.700 --> 00:15:57.200
good and everyone's happy, then these adapters are used by certain machines, and these ones

00:15:57.200 --> 00:16:00.200
are used by the other machines on the other port group.

00:16:00.200 --> 00:16:02.700
But what happens if I have a failure?

00:16:02.700 --> 00:16:06.200
What happens if NICs 4 or 5 were to fail?

00:16:06.200 --> 00:16:10.700
Well, what happened is NIC 6 could take over if one failed, and then 7 could take over

00:16:10.700 --> 00:16:15.200
if another one failed, and now all the traffic for both port groups are going over these

00:16:15.200 --> 00:16:17.700
two NICs, but traffic's still flowing.

00:16:17.700 --> 00:16:22.200
If they're all up, I have traffic divided across the different port groups, and I have

00:16:22.200 --> 00:16:29.700
this good even division of network traffic, but I don't have to survive a failure happen

00:16:29.700 --> 00:16:34.700
and all of a sudden VMs are going down if other physical adapters are still up.

00:16:34.700 --> 00:16:41.700
So, I can say okay to that, and now the port group settings are different right between

00:16:41.700 --> 00:16:42.700
those two.

00:16:42.700 --> 00:16:55.700
I can go to port group test 2 and edit its settings, and again, I can go to failover.

00:16:55.700 --> 00:16:59.700
I can override that, and what I could do is something completely different.

00:16:59.700 --> 00:17:08.199
I could say I want 7 to be standby, and I want 5 to be active.

00:17:08.699 --> 00:17:12.199
Whoops, did that again.

00:17:12.199 --> 00:17:13.199
Down, down.

00:17:13.199 --> 00:17:18.199
I want 5 to be active with 6, and I want 4 and 7 to be my standby.

00:17:18.199 --> 00:17:23.199
Now, here it's saying that 7 would be used first and then 4, but if I want that not to

00:17:23.199 --> 00:17:29.199
be so, I just move it down, or if I don't want 7 to ever back up on this particular

00:17:29.199 --> 00:17:32.199
port group for some reason, I can just call it unused.

00:17:32.199 --> 00:17:34.699
It's still claimed by the switch, right?

00:17:34.699 --> 00:17:39.699
The virtual switch is still using it, but it's not being used by this port group at

00:17:39.699 --> 00:17:40.699
all, okay?

00:17:41.699 --> 00:17:43.699
And I can say okay to that.

00:17:43.699 --> 00:17:50.699
So, now the only thing I've changed are the active adapters on this particular port group,

00:17:50.699 --> 00:17:51.699
right?

00:17:52.699 --> 00:17:58.699
I could set other options.

00:17:58.699 --> 00:18:02.699
For example, I could go to edit settings.

00:18:03.699 --> 00:18:05.699
I have my VLAN tag already set.

00:18:05.699 --> 00:18:10.699
Now, for security, do I want to allow promiscuous mode, which means all the traffic is going

00:18:10.699 --> 00:18:16.699
to be sent over to this port group, all layer 2 traffic.

00:18:17.699 --> 00:18:22.699
Right now, it's set to inherit it because override is not selected, so it's inheriting

00:18:22.699 --> 00:18:27.699
it from the switch, and the switch is saying to reject this, accept, accept, right?

00:18:27.699 --> 00:18:30.699
But I could go in and say MAC address changes.

00:18:30.699 --> 00:18:38.699
What MAC address changes means is that are we going to allow the virtual machine to change

00:18:38.699 --> 00:18:40.699
the MAC address of it?

00:18:40.699 --> 00:18:42.699
Can we use software to change this MAC address?

00:18:42.699 --> 00:18:47.699
Forge transmit means that I can actually send packets using MAC addresses other than the

00:18:47.699 --> 00:18:53.699
ones that are configured for my virtual machine in its settings in its BMX file, all right?

00:18:53.699 --> 00:18:54.699
All right.

00:18:54.699 --> 00:18:59.699
So, I could say, yeah, override that, and I want it to reject, reject.

00:18:59.699 --> 00:19:07.699
So, I have this locked down security for these particular – for this port group for these

00:19:07.699 --> 00:19:08.699
security settings.

00:19:08.699 --> 00:19:13.699
Traffic shaping, are we going to configure traffic shaping or not, right?

00:19:13.699 --> 00:19:16.699
Right now, it's currently disabled, but I could go in there and enable if I chose to,

00:19:16.699 --> 00:19:17.699
right?

00:19:17.699 --> 00:19:21.699
So, I can go in on a port group by port group level.

00:19:21.699 --> 00:19:26.699
I can choose what adapters I want them to use in what order for failover.

00:19:26.699 --> 00:19:32.699
I can configure traffic shaping on the port group level, security on the port group level,

00:19:32.699 --> 00:19:36.699
and the basic properties for, you know, for example, my VLAN tags, right?

00:19:36.699 --> 00:19:38.699
And I can say okay to that.

00:19:38.699 --> 00:19:43.699
So, this port group has different network adapter settings and security settings than

00:19:43.699 --> 00:19:49.699
these two, and these two have the same security settings as the switch, but different adapter

00:19:49.699 --> 00:19:51.699
settings than each other.

00:19:51.699 --> 00:19:53.699
This one is getting it from the switch.

00:19:53.699 --> 00:19:55.699
This guy inherited theirs, right?

00:19:55.699 --> 00:20:00.699
So, this is a quick little example of how virtual switches can be configured that way.

00:20:00.699 --> 00:20:05.699
The other thing we can quickly do is we can create a VMkernel port as well.

00:20:05.699 --> 00:20:11.699
So, this VMkernel port, I can go in and I'll say add networking, and I'll create a VMkernel

00:20:11.699 --> 00:20:15.699
adapter and select a target.

00:20:15.699 --> 00:20:21.699
I can say I want to select an existing switch, my switch one I have created here.

00:20:21.699 --> 00:20:23.699
Say next.

00:20:23.699 --> 00:20:27.699
The name of it, demo01.

00:20:27.699 --> 00:20:29.699
Do I want to add a VLAN to this?

00:20:29.699 --> 00:20:32.699
Yeah, I'll just use VLAN 230 for this one.

00:20:32.699 --> 00:20:35.699
Do I want to change the MTU?

00:20:35.699 --> 00:20:38.699
TCP-IP stack, right?

00:20:38.699 --> 00:20:42.699
What kind of options I want to set on here, right?

00:20:42.699 --> 00:20:43.699
I can create a custom one.

00:20:43.699 --> 00:20:45.699
I can create custom IP settings on this one.

00:20:45.699 --> 00:20:49.699
I can say am I using it for vMotion, for vSAN, for management?

00:20:49.699 --> 00:20:52.699
What am I using this for, right?

00:20:52.699 --> 00:20:55.699
Here, I'll just say I'm using it for vMotion, right?

00:20:55.699 --> 00:20:58.699
I think I already have one configured for that, so I'm going to create another one.

00:20:58.699 --> 00:21:03.699
Next, I'm going to my IPv4 settings, static.

00:21:03.699 --> 00:21:16.699
I'm just going to say 10.10.10.11, 255.255.255.0, my gateway.

00:21:16.699 --> 00:21:20.699
Now, if I don't select this right here, it will use the default gateway that's

00:21:20.699 --> 00:21:24.699
configured by default on my original VMkernel port or it has a gateway on there,

00:21:24.699 --> 00:21:27.699
or I could give it another gateway that's on this network here.

00:21:27.699 --> 00:21:29.699
I'm going to just go ahead and leave it the way it is so that way

00:21:29.699 --> 00:21:32.699
router check would go through that interface.

00:21:32.699 --> 00:21:35.699
Because for my vMotion network, it's not going to go outside this network anyway.

00:21:35.699 --> 00:21:39.699
Now, if I was doing, like, metro, like, long-range vMotions and things like that,

00:21:39.699 --> 00:21:43.699
I'd probably put a gateway on there.

00:21:43.699 --> 00:21:45.699
And finish.

00:21:45.699 --> 00:21:46.699
And this is going to go through.

00:21:46.699 --> 00:21:48.699
It went through actually really quick.

00:21:48.699 --> 00:21:51.699
If I wanted to, like, see what's going on, I could see here, oh, look,

00:21:51.699 --> 00:21:56.699
except that virtual NIC, no problem.

00:21:57.699 --> 00:22:01.699
Got my VMkernel port configured on my virtual switch, on my virtual switch here.

00:22:01.699 --> 00:22:05.699
I have my port groups, and here's my new VMkernel demo.

00:22:05.699 --> 00:22:07.699
So, I've created a VMkernel port.

00:22:07.699 --> 00:22:09.699
I've created a virtual switch.

00:22:09.699 --> 00:22:11.699
I've managed my physical adapters.

00:22:11.699 --> 00:22:12.699
Okay.

00:22:12.699 --> 00:22:14.699
I've pretty much done everything that we can, you know,

00:22:14.699 --> 00:22:17.699
do here on these standard virtual switches.

00:22:17.699 --> 00:22:20.699
Not a lot of settings, but there's plenty of settings, you know,

00:22:20.699 --> 00:22:23.699
for a standard switch, a lot to remember.

00:22:23.699 --> 00:22:25.699
So, that was a standard virtual switch.

00:22:25.699 --> 00:22:27.699
Next, we're going to discuss distributed switches

00:22:27.699 --> 00:22:29.699
or distributed virtual switches.

00:22:29.699 --> 00:22:34.699
A distributed virtual switch is not created on a host-by-host basis.

00:22:34.699 --> 00:22:38.699
We create a distributed virtual switch on our data center,

00:22:38.699 --> 00:22:41.699
and then we can assign it to hosts within that data center.

00:22:41.699 --> 00:22:47.699
And from there, we can assign options to that distributed virtual switch.

00:22:47.699 --> 00:22:49.699
We can assign a VLAN tag.

00:22:49.699 --> 00:22:51.699
We can do all the same things we did on a standard switch.

00:22:51.699 --> 00:22:52.699
We can set our security.

00:22:52.699 --> 00:22:55.699
We can set our load balancing and our failover policies

00:22:55.699 --> 00:22:58.699
and our MTU size and all that kind of stuff.

00:22:58.699 --> 00:23:02.699
We can also create port groups on a distributed virtual switch,

00:23:02.699 --> 00:23:06.699
and we can assign different adapter settings, different failover,

00:23:06.699 --> 00:23:10.699
different active standby, different security settings.

00:23:10.699 --> 00:23:13.699
You'll see there's also more advanced options we'll get into

00:23:13.699 --> 00:23:15.699
for distributed virtual switches as well.

00:23:15.699 --> 00:23:18.699
And we can set all that up on the distributed virtual switch.

00:23:18.699 --> 00:23:21.699
Then we can assign it to a host.

00:23:21.699 --> 00:23:24.699
Now, I might say on my distributed virtual switch

00:23:24.699 --> 00:23:27.699
that I want it to have four uplinks.

00:23:27.699 --> 00:23:30.699
And we don't assign physical adapters to that distributed virtual switch.

00:23:30.699 --> 00:23:34.699
When we create it, we say how many uplinks we want it to have.

00:23:34.699 --> 00:23:39.699
Then when we assign it to a host, we say on that host that we want you

00:23:39.699 --> 00:23:42.699
to use adapters 1, 2, 3, and 4.

00:23:42.699 --> 00:23:45.699
Or we can say 4, 5, 6, and 7, or whatever it is.

00:23:45.699 --> 00:23:47.699
We can choose whatever adapters.

00:23:47.699 --> 00:23:51.699
And we can say I want adapter 4 to go to uplink 1,

00:23:51.699 --> 00:23:53.699
and adapter 5 to go to uplink 2,

00:23:53.699 --> 00:23:58.699
and adapter 6 to go to uplink 3, and things like that, obviously.

00:23:58.699 --> 00:24:04.699
And so now I have assigned physical adapters to the uplinks

00:24:04.699 --> 00:24:06.699
on that distributed virtual switch.

00:24:06.699 --> 00:24:09.699
I said what NIC goes to what uplink.

00:24:09.699 --> 00:24:12.699
And the reason why you might want to do that is maybe for some reason

00:24:12.699 --> 00:24:17.699
this is not probable, but maybe like adapter 4 is a 1 gig,

00:24:17.699 --> 00:24:18.699
and the other three are 10 gigs.

00:24:18.699 --> 00:24:21.699
So I want adapter 4 to always go to uplink 1,

00:24:21.699 --> 00:24:27.699
because uplink 1 is going to go to a VM kernel port that is just going to

00:24:27.699 --> 00:24:30.699
use that 1 gig connection, and it's just for simple management,

00:24:30.699 --> 00:24:32.699
and that 1 gig connection is fine.

00:24:32.699 --> 00:24:36.699
And then the other three uplinks are my 10 gigs, or 25 gigs.

00:24:36.699 --> 00:24:39.699
25 is getting really common now on your ESXi hosts,

00:24:39.699 --> 00:24:41.699
are these 25 gig uplinks.

00:24:41.699 --> 00:24:46.699
And so I want to know which one is my 1 gig and which ones are my 25 gigs,

00:24:46.699 --> 00:24:49.699
because that uplink is going to only be attached to that VM kernel port,

00:24:49.699 --> 00:24:51.699
and then the other three are going to be for VMs

00:24:51.699 --> 00:24:53.699
or for whatever you're going to be using them for, right?

00:24:53.699 --> 00:24:59.699
So by knowing what physical NIC on the host I've assigned to the uplink,

00:24:59.699 --> 00:25:02.699
I now know my configuration will be valid across all of them.

00:25:02.699 --> 00:25:06.699
If I were to say just these four NICs and they land wherever they happen

00:25:06.699 --> 00:25:10.699
to land, then maybe my 25 gig is going to that VM kernel port,

00:25:10.699 --> 00:25:12.699
and that 1 gig is used for VMware traffic or something like that.

00:25:12.699 --> 00:25:14.699
Obviously, that would be a horrible thing.

00:25:14.699 --> 00:25:18.699
So by being able to assign an adapter to a specific uplink,

00:25:18.699 --> 00:25:24.699
and then that specific uplink is going to be used for my adapter policies,

00:25:24.699 --> 00:25:28.699
for my load balancing and my failover order and all that kind of stuff,

00:25:28.699 --> 00:25:32.699
I now know that my configuration is going to be valid

00:25:32.699 --> 00:25:35.699
across my infrastructure, okay?

00:25:35.699 --> 00:25:38.699
And I don't have to reconfigure it again over and over and over

00:25:38.699 --> 00:25:41.699
and over and over across 25, 30, 40 hosts.

00:25:41.699 --> 00:25:45.699
I only have to configure it once and then assign it to the host,

00:25:45.699 --> 00:25:49.699
choose my adapters to my uplink mapping, and I'm done.

00:25:49.699 --> 00:25:51.699
So let's go ahead and take a look at this.

00:25:51.699 --> 00:25:54.699
So distributed virtual switches created off the data center level.

00:25:54.699 --> 00:25:57.699
We go to our vSphere client, and we're going to go into there,

00:25:57.699 --> 00:26:01.699
and we're going to build the distributed virtual switch,

00:26:01.699 --> 00:26:03.699
and we're going to map it to our host.

00:26:03.699 --> 00:26:06.699
We allow folks to create configurations that are going to be assigned

00:26:06.699 --> 00:26:10.699
to multiple hosts and not have to reconfigure them every single time

00:26:10.699 --> 00:26:13.699
and worry about things like typos and things like that.

00:26:13.699 --> 00:26:17.699
One point for management, if down the road, let's say in a year,

00:26:17.699 --> 00:26:21.699
all of a sudden I need another VLAN in my vSphere environment,

00:26:21.699 --> 00:26:23.699
I can go to my distributed virtual switch,

00:26:23.699 --> 00:26:27.699
I can create a new port group on that switch, assign it a new VLAN,

00:26:27.699 --> 00:26:32.699
and all my hosts are going to get that new port group with that new VLAN.

00:26:32.699 --> 00:26:34.699
Then I need to edit my virtual machines

00:26:34.699 --> 00:26:37.699
and tell them to use that new port group that I just now pushed out,

00:26:37.699 --> 00:26:41.699
or maybe it's not changing VNs, maybe I'm going to add a new group of VNs.

00:26:41.699 --> 00:26:45.699
I'm deploying a new security appliance that needs its own VLAN.

00:26:45.699 --> 00:26:50.699
So I'm going to map it to that when I deploy it.

00:26:50.699 --> 00:26:53.699
So here we're seeing the distributed virtual switch architecture.

00:26:53.699 --> 00:26:55.699
What we're seeing on the far left,

00:26:55.699 --> 00:26:59.699
we have a host that at the top has a distributed virtual switch,

00:26:59.699 --> 00:27:02.699
and below that it has a regular standard virtual switch.

00:27:02.699 --> 00:27:04.699
We call it a host switch here.

00:27:04.699 --> 00:27:08.699
And that distributed virtual switch is going to be mapped to a certain adapter

00:27:08.699 --> 00:27:11.699
on the host that goes to the switch.

00:27:11.699 --> 00:27:16.699
Now in the middle we also see a host that has a host switch only.

00:27:16.699 --> 00:27:21.699
Now that host with a host switch only is not using that distributed virtual switch.

00:27:21.699 --> 00:27:25.699
And then the last host only has a distributed virtual switch.

00:27:25.699 --> 00:27:30.699
So you can see you can use distributed virtual switches exclusively on a host

00:27:30.699 --> 00:27:37.699
or you can mix them with standard virtual switches on a host.

00:27:37.699 --> 00:27:42.699
So the created distributed virtual switch is basically pretty easy to do,

00:27:42.699 --> 00:27:44.699
but there are a lot of questions,

00:27:44.699 --> 00:27:47.699
and you want to be really careful when you do it that you're basically the

00:27:47.699 --> 00:27:50.699
biggest challenge or the thing you need to be most careful about,

00:27:50.699 --> 00:27:52.699
it's not really a challenge,

00:27:52.699 --> 00:27:54.699
is choosing the host when you apply it to the host,

00:27:54.699 --> 00:27:56.699
making sure you choose the right adapters.

00:27:56.699 --> 00:27:59.699
You don't want to choose the wrong adapters or the wrong configuration

00:27:59.699 --> 00:28:01.699
or something like that.

00:28:01.699 --> 00:28:03.699
You want to make sure you get this all configured correctly.

00:28:03.699 --> 00:28:06.699
So here we can go into our data center,

00:28:06.699 --> 00:28:08.699
distribute switches and create a new distributed switch.

00:28:08.699 --> 00:28:10.699
We're going to give it a name.

00:28:10.699 --> 00:28:12.699
We're going to set a location for the distributed virtual switch,

00:28:12.699 --> 00:28:15.699
and we'll see there's a bunch of options we can start configuring.

00:28:15.699 --> 00:28:19.699
So when you are creating new distributed virtual switches,

00:28:19.699 --> 00:28:23.699
this is not software that gets installed on a host.

00:28:23.699 --> 00:28:27.699
It's integrated into the ESXi operating system.

00:28:27.699 --> 00:28:30.699
And because it is integrated with the ESXi operating system,

00:28:30.699 --> 00:28:35.699
different versions of ESXi may have different versions of distributed

00:28:35.699 --> 00:28:36.699
virtual switches.

00:28:36.699 --> 00:28:38.699
So if when I create this,

00:28:38.699 --> 00:28:42.699
I say I want to create a distributed virtual switch that's compatible with

00:28:42.699 --> 00:28:47.699
ESXi 8 and I have hosts that are running 7 or, you know,

00:28:47.699 --> 00:28:50.699
some older version, 7 update 3 or, you know, 6.5 maybe,

00:28:51.699 --> 00:28:54.699
that distributed virtual switch won't be compatible with those older

00:28:54.699 --> 00:28:56.699
versions of ESXi.

00:28:56.699 --> 00:29:00.699
If I know my entire infrastructure or at least everything in this data

00:29:00.699 --> 00:29:04.699
center is all going to be 8, then I can choose 8.

00:29:04.699 --> 00:29:07.699
But if I know I have to be compatible with some older hosts,

00:29:07.699 --> 00:29:09.699
I would need to choose a version that's compatible with that older

00:29:09.699 --> 00:29:10.699
host.

00:29:10.699 --> 00:29:13.699
And then that older version might have limitations on features or,

00:29:13.699 --> 00:29:16.699
you know, the size it can grow to and things like that.

00:29:16.699 --> 00:29:17.699
Right?

00:29:17.699 --> 00:29:19.699
So I'm going to choose a version.

00:29:19.699 --> 00:29:21.699
Then I'm going to say how many ports do I want it to use?

00:29:21.699 --> 00:29:28.699
Now it's possible that we want it to be a set physical size and it's

00:29:28.699 --> 00:29:33.699
going to be, you know, 48 ports or whatever we're going to choose,

00:29:33.699 --> 00:29:34.699
a number.

00:29:34.699 --> 00:29:37.699
We can also set it to be an initial size of, say,

00:29:37.699 --> 00:29:39.699
let's say, for example, 8.

00:29:39.699 --> 00:29:42.699
But we want to be able to grow as needed.

00:29:42.699 --> 00:29:45.699
If you choose a switch that can, you know,

00:29:46.699 --> 00:29:49.699
expand and grow as needed and you set a number of ports on it,

00:29:49.699 --> 00:29:53.699
say, for example, 8, every time it grows, it'll grow by 8.

00:29:53.699 --> 00:29:57.699
If I set 24, every time it grows, it grows by 24.

00:29:57.699 --> 00:30:01.699
So you probably want to choose a relatively smallish number, like,

00:30:01.699 --> 00:30:03.699
you know, 8 is probably a pretty decent number, 12,

00:30:03.699 --> 00:30:05.699
something like that, so that when it grows,

00:30:05.699 --> 00:30:07.699
it doesn't grow in huge increments.

00:30:07.699 --> 00:30:10.699
And I only needed one more port and I got 24 more ports.

00:30:10.699 --> 00:30:11.699
I got 48 more ports.

00:30:11.699 --> 00:30:13.699
I only needed one and I got 12.

00:30:13.699 --> 00:30:16.699
So that gives me some room to add some more VMs to this switch on this

00:30:16.699 --> 00:30:18.699
host as needed, right?

00:30:18.699 --> 00:30:21.699
So how many ports do we want to use?

00:30:21.699 --> 00:30:24.699
Are we going to enable I-O control?

00:30:24.699 --> 00:30:29.699
So basically I-O control will allow us to prioritize traffic,

00:30:29.699 --> 00:30:33.699
and we're going to say basically that certain port groups can't overrun

00:30:33.699 --> 00:30:35.699
other port groups.

00:30:35.699 --> 00:30:38.699
We can, you know, set up how much I-O is going to be allowed to be used.

00:30:38.699 --> 00:30:40.699
We can put a limit on there.

00:30:40.699 --> 00:30:43.699
We can say that we want to basically do it based on shares,

00:30:43.699 --> 00:30:46.699
and we can say that as long as there's no contention,

00:30:46.699 --> 00:30:47.699
use what you want.

00:30:47.699 --> 00:30:49.699
But if there is contention, it's divided 50-50.

00:30:49.699 --> 00:30:52.699
Or if there's contention, this one gets 60 and that one gets 40,

00:30:52.699 --> 00:30:53.699
things like that.

00:30:53.699 --> 00:30:57.699
You can set basically the amount of access that can be used for I-O.

00:30:57.699 --> 00:31:00.699
And you'll probably want to create a default port group, right?

00:31:00.699 --> 00:31:02.699
And, of course, port groups require a name.

00:31:02.699 --> 00:31:05.699
So location of the distributed virtual switch,

00:31:05.699 --> 00:31:07.699
the distributed virtual switch port groups,

00:31:07.699 --> 00:31:09.699
the name of the port group,

00:31:09.699 --> 00:31:12.699
and you're going to select the distributed virtual switch to put the port

00:31:12.699 --> 00:31:16.699
group on when you create that port group.

00:31:16.699 --> 00:31:21.699
So set bindings, static bindings of port to virtual machine mapping.

00:31:21.699 --> 00:31:26.699
So what that means is that when a VM is connecting to the virtual switch,

00:31:26.699 --> 00:31:30.699
if it gets connected to port 8, it's always going to use port 8.

00:31:30.699 --> 00:31:33.699
If you shut down the VM and you restart it up a week later,

00:31:33.699 --> 00:31:34.699
it goes to port 8.

00:31:34.699 --> 00:31:36.699
Ephemeral means there's no binding.

00:31:36.699 --> 00:31:38.699
So when you start it up, it goes to port 8.

00:31:38.699 --> 00:31:40.699
You shut it down, another machine takes port 8.

00:31:40.699 --> 00:31:42.699
And when it starts back up again, it says, oh, I'm going to take port 6

00:31:42.699 --> 00:31:44.699
or I'm going to take port 9, whatever's available.

00:31:44.699 --> 00:31:47.699
So ephemeral, there's no binding from VM to port,

00:31:47.699 --> 00:31:49.699
and with static bindings,

00:31:49.699 --> 00:31:52.699
that VM is going to always attach to the same port on the virtual switch.

00:31:52.699 --> 00:31:58.699
So port allocation, elastic means there's no set certain size for the

00:31:58.699 --> 00:31:59.699
virtual switch.

00:31:59.699 --> 00:32:02.699
That's when we're talking about if you said the switch, you know,

00:32:02.699 --> 00:32:05.699
it's going to be elastic and we're going to create it with 24 ports.

00:32:05.699 --> 00:32:08.699
When it fills up the 24 ports, it adds 24 more.

00:32:08.699 --> 00:32:12.699
If I had said 8, when it fills up 8, it'll add 8 more.

00:32:12.699 --> 00:32:14.699
So you probably want to choose a smaller number

00:32:14.699 --> 00:32:17.699
and let it grow in smaller increments, but not ridiculous.

00:32:17.699 --> 00:32:19.699
You don't want to make it too small where it's growing constantly.

00:32:19.699 --> 00:32:22.699
You want it to grow occasionally when needed.

00:32:22.699 --> 00:32:26.699
So 8 or 12 or something like that might be a good start.

00:32:26.699 --> 00:32:28.699
In the beginning, it might grow several times,

00:32:28.699 --> 00:32:30.699
but then it'll kind of settle into a rhythm

00:32:30.699 --> 00:32:32.699
and you won't have to keep growing it all the time.

00:32:32.699 --> 00:32:35.699
Fixed means it's 48 ports, always going to be 48 ports,

00:32:35.699 --> 00:32:37.699
it'll never be more than 48 ports, right,

00:32:37.699 --> 00:32:41.699
or 24 or 80 or whatever number you choose.

00:32:44.699 --> 00:32:46.699
So how many ports do you need, right?

00:32:46.699 --> 00:32:48.699
That's one that we just discussed.

00:32:48.699 --> 00:32:51.699
Network resource pool, you can assign a network resource pool

00:32:51.699 --> 00:32:55.699
to the port group or distributed virtual switch.

00:32:55.699 --> 00:32:59.699
So basically when you assign it to a port group on a distributed virtual

00:32:59.699 --> 00:33:04.699
switch, you can specify a maximum amount of bandwidth it uses.

00:33:04.699 --> 00:33:07.699
You can say it's unlimited, you can use as much as you need,

00:33:07.699 --> 00:33:11.699
or you can put a cap on there and say the maximum it can use is

00:33:11.699 --> 00:33:13.699
1,000 megabits, right?

00:33:13.699 --> 00:33:15.699
If you've got a 10-gig NIC and you give it 1,000 megabits,

00:33:15.699 --> 00:33:20.699
it's got a gig, but it can't exceed that ever, all right?

00:33:20.699 --> 00:33:23.699
The other things on your resource pool you can say are shares.

00:33:23.699 --> 00:33:27.699
So you can say this resource pool has got 100 shares

00:33:28.699 --> 00:33:30.699
but its usage is unlimited.

00:33:30.699 --> 00:33:34.699
What that means is it can use all it wants if there's no contention for

00:33:34.699 --> 00:33:35.699
the resource.

00:33:35.699 --> 00:33:37.699
If there's contention for the resource,

00:33:37.699 --> 00:33:40.699
what's going to happen is they'll look at the shares and say, oh,

00:33:40.699 --> 00:33:42.699
100 and 100, so it's 50-50.

00:33:42.699 --> 00:33:48.699
If one port group has 10 VMs in it, it gets half the bandwidth.

00:33:48.699 --> 00:33:51.699
If the other port group has 30 VMs in it, three times more,

00:33:51.699 --> 00:33:53.699
but the same number of shares,

00:33:53.699 --> 00:33:56.699
those 30 VMs have the same amount of bandwidth that share amongst them,

00:33:56.699 --> 00:33:58.699
as those 10 have, right?

00:33:58.699 --> 00:34:00.699
So obviously the 10 have a lot more bandwidth.

00:34:00.699 --> 00:34:07.699
You could also go in and say that the one with 30 VMs has 300 shares,

00:34:07.699 --> 00:34:10.699
and then the one with 10 VMs only has 100 shares, right?

00:34:10.699 --> 00:34:14.699
So that way the VMs would kind of be more equal.

00:34:14.699 --> 00:34:16.699
But then, of course, if you take away VMs or add VMs,

00:34:16.699 --> 00:34:18.699
it starts to change, the map changes again.

00:34:18.699 --> 00:34:22.699
But you can assign a port group more shares,

00:34:22.699 --> 00:34:24.699
giving it access to more bandwidth.

00:34:24.699 --> 00:34:27.699
Now, that bandwidth is only up to its cap.

00:34:27.699 --> 00:34:36.699
So if you put a limit on there of, say, you know, 5,000 megabits a second,

00:34:36.699 --> 00:34:40.699
when it hits that 5,000, it will never be able to exceed it,

00:34:40.699 --> 00:34:43.699
even if it's available, even if there's no contention.

00:34:43.699 --> 00:34:45.699
No one else is using that adapter for anything.

00:34:45.699 --> 00:34:48.699
There's 5,000 megabits still available.

00:34:48.699 --> 00:34:51.699
When you hit that limit, it's a limit you can't exceed it.

00:34:51.699 --> 00:34:53.699
So just be careful of that.

00:34:54.699 --> 00:34:55.699
Limits are useful.

00:34:55.699 --> 00:34:58.699
There's reasons to use them, but you don't use them all the time

00:34:58.699 --> 00:35:02.699
because you don't want to be wasting resources that could be used for something.

00:35:02.699 --> 00:35:05.699
An example when you probably would use something like a limit is,

00:35:05.699 --> 00:35:09.699
let's say you're a cloud provider and you're reselling a resource,

00:35:09.699 --> 00:35:13.699
and a customer is paying for a certain amount of bandwidth.

00:35:13.699 --> 00:35:16.699
You put a cap on there, you put a limit on there,

00:35:16.699 --> 00:35:19.699
and don't let them exceed that because they're not paying for more than that.

00:35:19.699 --> 00:35:20.699
It's available.

00:35:20.699 --> 00:35:21.699
You have it available to them.

00:35:21.699 --> 00:35:25.699
But if you let them use it and then later on you take it away because now you

00:35:25.699 --> 00:35:29.699
added more customers and so now you get to enforce your limits,

00:35:29.699 --> 00:35:32.699
you add a limit on there, now they're going to see their performance

00:35:32.699 --> 00:35:35.699
degradate and they're going to complain.

00:35:35.699 --> 00:35:38.699
And also you're hurting your own revenue because they're not paying for more

00:35:38.699 --> 00:35:40.699
than that, so why would you give them more than that?

00:35:40.699 --> 00:35:42.699
If they need more, they should pay you for more.

00:35:42.699 --> 00:35:46.699
So VLAN types, you can say there's no VLAN configured.

00:35:46.699 --> 00:35:48.699
We are going to add a VLAN, so then, of course,

00:35:48.699 --> 00:35:50.699
it's going to put a VLAN tag on there.

00:35:50.699 --> 00:35:54.699
VLAN trunking is exactly like what it would be if you were on a switch.

00:35:54.699 --> 00:35:58.699
So maybe on a switch you have the physical interface and the switch says,

00:35:58.699 --> 00:36:02.699
I'm going to allow VLANs, you know, 10, 20, 30, and 40.

00:36:02.699 --> 00:36:06.699
On a port group, you might just say this port group is VLAN 20

00:36:06.699 --> 00:36:08.699
and the next one is VLAN 30.

00:36:08.699 --> 00:36:13.699
But the next one might be 40 and 50 or 40 through 50 or something like that.

00:36:13.699 --> 00:36:17.699
Now, of course, 40 through 50 would have to be enabled on your switch.

00:36:17.699 --> 00:36:21.699
And the reason why you might do that on a port group is because you might have

00:36:21.699 --> 00:36:25.699
VMs in there that are able to assign their own VLAN tags.

00:36:25.699 --> 00:36:27.699
Like a Windows VM doesn't usually do that.

00:36:27.699 --> 00:36:29.699
Linux VMs don't usually do that.

00:36:29.699 --> 00:36:34.699
Macs aren't usually VMs, but they don't usually do that either.

00:36:34.699 --> 00:36:37.699
But certain types of appliances might be able to do that.

00:36:37.699 --> 00:36:40.699
They might be configured to assign their own VLAN tag.

00:36:40.699 --> 00:36:45.699
And if they are, then you can just put them into a port group if you want them

00:36:45.699 --> 00:36:48.699
to share that, and you can trunk multiple VLANs.

00:36:48.699 --> 00:36:49.699
Let me use an application.

00:36:49.699 --> 00:36:53.699
And one part of the application uses VLAN 31.

00:36:53.699 --> 00:36:56.699
The other part of the application is on VLAN 32.

00:36:56.699 --> 00:36:58.699
You can put them on the same port group,

00:36:58.699 --> 00:37:04.699
and then you can trunk both those VLANs to that port group.

00:37:04.699 --> 00:37:07.699
A private VLAN is a little bit kind of different.

00:37:07.699 --> 00:37:11.699
So a private VLAN is that there's ports on your virtual switch,

00:37:11.699 --> 00:37:16.699
and a device on port one cannot talk to a device on port two

00:37:16.699 --> 00:37:20.699
or a device on port three on that same private VLAN.

00:37:20.699 --> 00:37:25.699
They can only talk to what would be an uplink port or a link port.

00:37:25.699 --> 00:37:29.699
That port might then connect out to the Internet only.

00:37:29.699 --> 00:37:33.699
So what you just did is you created a bunch of a network

00:37:33.699 --> 00:37:35.699
with a bunch of machines on there.

00:37:35.699 --> 00:37:37.699
It would be virtual machines or if this is in your physical environment,

00:37:37.699 --> 00:37:39.699
physical, you know, machines on a private VLAN.

00:37:40.699 --> 00:37:43.699
In this case, virtual machines that are all on the same VLAN,

00:37:43.699 --> 00:37:49.699
on the same virtual switch, that can only talk to out to the Internet.

00:37:49.699 --> 00:37:51.699
They cannot talk and see each other.

00:37:51.699 --> 00:37:52.699
All right?

00:37:52.699 --> 00:37:54.699
So if you want to create some kind of security rule like that,

00:37:54.699 --> 00:37:56.699
you could do that with a private VLAN.

00:37:56.699 --> 00:37:59.699
It's kind of like what you see a lot of times on Wi-Fi, on guest networks,

00:37:59.699 --> 00:38:02.699
where the guest network will let, you know,

00:38:02.699 --> 00:38:05.699
your phone connect to the Wi-Fi and then get up to the Internet,

00:38:05.699 --> 00:38:08.699
but you can't see the guy sitting next to you who's on the same Wi-Fi,

00:38:08.699 --> 00:38:09.699
on the same SSID.

00:38:09.699 --> 00:38:13.699
It's kind of like that.

00:38:13.699 --> 00:38:14.699
So security for port groups.

00:38:14.699 --> 00:38:17.699
We have promiscuous mode, MAC address changes, and forged transmit.

00:38:17.699 --> 00:38:20.699
Promiscuous mode means that if you have set to accept,

00:38:20.699 --> 00:38:23.699
that it's going to allow all the different traffic,

00:38:23.699 --> 00:38:27.699
even if it's not destined to your MAC address, to come in on that port.

00:38:27.699 --> 00:38:29.699
So if you were running some kind of a network monitoring tool

00:38:29.699 --> 00:38:30.699
or something like that,

00:38:30.699 --> 00:38:32.699
you would probably enable promiscuous mode on there.

00:38:32.699 --> 00:38:34.699
If you have set to reject, then that's rejected.

00:38:35.699 --> 00:38:40.699
So MAC address changes means the guest OS can change its MAC address,

00:38:40.699 --> 00:38:41.699
right,

00:38:41.699 --> 00:38:46.699
and send out packets on the network with a different MAC address than what

00:38:46.699 --> 00:38:47.699
it's configured for.

00:38:47.699 --> 00:38:49.699
Reject means that it can't.

00:38:49.699 --> 00:38:53.699
Forged transmit means all outbound frames are allowed by the switch,

00:38:53.699 --> 00:38:55.699
regardless of what MAC address they're using.

00:38:55.699 --> 00:38:58.699
So they can actually use a whole bunch of different MAC addresses.

00:38:58.699 --> 00:39:01.699
They're forging a whole bunch of different MACs and sending them out,

00:39:01.699 --> 00:39:04.699
whereas the guest OS is saying, this is my MAC address.

00:39:04.699 --> 00:39:08.699
But the guest OS MAC address is different than what's in the VMX file.

00:39:08.699 --> 00:39:11.699
That's the configuration file for that VM.

00:39:11.699 --> 00:39:12.699
The VMX file says one thing.

00:39:12.699 --> 00:39:15.699
The guest OS is saying, nope, my MAC address is this.

00:39:15.699 --> 00:39:17.699
That's a MAC address change.

00:39:17.699 --> 00:39:22.699
Forged transmit means that it's just putting in different MAC addresses,

00:39:22.699 --> 00:39:24.699
the application is, and sending those out.

00:39:24.699 --> 00:39:28.699
And it could be the same one over and over or a bunch of different ones.

00:39:31.699 --> 00:39:33.699
Traffic shaping.

00:39:33.699 --> 00:39:37.699
So basically what you're doing with traffic shaping is you're saying how

00:39:37.699 --> 00:39:39.699
much traffic you're allowed to use.

00:39:39.699 --> 00:39:41.699
You can say the average over time needs to be this.

00:39:41.699 --> 00:39:45.699
You're allowed to peak, right, because here's the most that you can use.

00:39:45.699 --> 00:39:49.699
A burst means you can go even higher, but it can't exceed the average.

00:39:49.699 --> 00:39:55.699
So you can say the average is this, but the peak could be this.

00:39:55.699 --> 00:39:57.699
A burst could be even higher.

00:39:57.699 --> 00:40:02.699
But none of these can exceed what your average is, right?

00:40:02.699 --> 00:40:05.699
So if your average has to be 1,000 megabits,

00:40:05.699 --> 00:40:08.699
maybe your peak would be 1,500.

00:40:08.699 --> 00:40:13.699
But you can't peak to 1,500 for so long that you exceed your average over

00:40:13.699 --> 00:40:18.699
time, and you can't burst, which means like a spike and back down, again,

00:40:18.699 --> 00:40:20.699
that would exceed your average.

00:40:20.699 --> 00:40:23.699
So all those have to be enforced, all right?

00:40:23.699 --> 00:40:26.699
So each port receives traffic shaping policies individually.

00:40:26.699 --> 00:40:29.699
What that means is that when you set a traffic shaping policy,

00:40:29.699 --> 00:40:30.699
it's on a port-by-port basis.

00:40:30.699 --> 00:40:32.699
That is assigned to the port.

00:40:32.699 --> 00:40:34.699
So you're saying what the average, what the peak,

00:40:34.699 --> 00:40:37.699
and what the burst is for a VM on a port,

00:40:37.699 --> 00:40:42.699
not for all the VMs on the port group.

00:40:42.699 --> 00:40:45.699
So a 10-minute failover, we discussed this actually quite a bit already,

00:40:45.699 --> 00:40:50.699
right, but it basically allows the traffic to be shared across the adapters

00:40:50.699 --> 00:40:53.699
that are assigned to the virtual switch or to the port group.

00:40:53.699 --> 00:40:55.699
What happens if an adapter fails?

00:40:55.699 --> 00:40:56.699
Is there a failover order?

00:40:56.699 --> 00:40:58.699
Do you have ones reserved for failover?

00:40:58.699 --> 00:41:02.699
And so how do you configure this?

00:41:02.699 --> 00:41:05.699
It's if you're teaming in your failover options, right?

00:41:05.699 --> 00:41:06.699
So load balancing.

00:41:06.699 --> 00:41:08.699
So your load balancing is saying,

00:41:08.699 --> 00:41:12.699
if I have more than one adapter selected as my active adapter,

00:41:12.699 --> 00:41:15.699
how is traffic divided across those adapters?

00:41:15.699 --> 00:41:19.699
First of all, a VM is only going to use one adapter.

00:41:20.699 --> 00:41:24.699
It's not going to send traffic from a single VM across multiple adapters.

00:41:24.699 --> 00:41:26.699
It's only going to use one adapter.

00:41:26.699 --> 00:41:30.699
So how do I know what adapter it's going to use?

00:41:30.699 --> 00:41:31.699
So you have different options.

00:41:31.699 --> 00:41:35.699
You have an IP hash where it just hashes the IP address of the VM,

00:41:35.699 --> 00:41:38.699
assigns it to a port.

00:41:38.699 --> 00:41:42.699
There is a source MAC address hash, right?

00:41:42.699 --> 00:41:44.699
Physical NIC load,

00:41:44.699 --> 00:41:48.699
that actually is going to only be done on a distributed virtual switch.

00:41:48.699 --> 00:41:51.699
That is not an option on a standard switch, okay?

00:41:51.699 --> 00:41:53.699
And then you also have your explicit failover.

00:41:53.699 --> 00:41:58.699
So if one adapter were to fail, which one becomes next in that load?

00:41:58.699 --> 00:42:02.699
And then the hash has to be redone to use that new adapter.

00:42:02.699 --> 00:42:03.699
You wouldn't have to redo the hash, actually.

00:42:03.699 --> 00:42:07.699
Just all the ones that were going to the failed adapter would go to the

00:42:07.699 --> 00:42:08.699
new replacement adapter.

00:42:08.699 --> 00:42:10.699
So I slightly misspoke there.

00:42:10.699 --> 00:42:12.699
Failure detection.

00:42:12.699 --> 00:42:14.699
How do we know if an adapter fails?

00:42:14.699 --> 00:42:15.699
How do we know an adapter fails?

00:42:15.699 --> 00:42:19.699
And I need to take that one out and put a standby in its place.

00:42:19.699 --> 00:42:21.699
There's two different options.

00:42:21.699 --> 00:42:24.699
One is link status, which means if the adapter goes dark,

00:42:24.699 --> 00:42:28.699
if the little link light on there were to go out, that adapter is down.

00:42:28.699 --> 00:42:34.699
But what happens if, for example, the switch it's connected to shows up,

00:42:34.699 --> 00:42:37.699
but it can't connect to anything else past that switch?

00:42:37.699 --> 00:42:39.699
You could use beaconing for that.

00:42:39.699 --> 00:42:42.699
So if this adapter can still talk to it, you know,

00:42:43.699 --> 00:42:47.699
the adapter is able to get sent out traffic and it's seen by other hosts

00:42:47.699 --> 00:42:50.699
or by other adapters, then everything's good.

00:42:50.699 --> 00:42:53.699
But if I can't see anything else, I see a link light on,

00:42:53.699 --> 00:42:58.699
but I can't see any traffic from my beacons, then I consider it down.

00:42:58.699 --> 00:43:00.699
Most people do link status.

00:43:00.699 --> 00:43:01.699
They don't do the beaconing.

00:43:01.699 --> 00:43:03.699
They just do link status.

00:43:03.699 --> 00:43:05.699
If you're doing an IP hash, then you don't use beaconing.

00:43:05.699 --> 00:43:10.699
You just use link status.

00:43:10.699 --> 00:43:14.699
So what happens if a adapter were to fail?

00:43:14.699 --> 00:43:16.699
So I have my link status.

00:43:16.699 --> 00:43:17.699
The link light goes out.

00:43:17.699 --> 00:43:18.699
The adapter goes down.

00:43:18.699 --> 00:43:20.699
I see the adapter status is down.

00:43:20.699 --> 00:43:22.699
I take it out of my active.

00:43:22.699 --> 00:43:25.699
I put a standby adapter in its place.

00:43:25.699 --> 00:43:26.699
What do I do now?

00:43:26.699 --> 00:43:29.699
Well, most of the time, you're going to want to say you want to notify

00:43:29.699 --> 00:43:30.699
your switches.

00:43:30.699 --> 00:43:32.699
You want to check the box that says notify switches.

00:43:32.699 --> 00:43:35.699
Because what will happen is by notifying the switches,

00:43:35.699 --> 00:43:39.699
what it does is it sends a packet out with the MAC address of all the VMs

00:43:39.699 --> 00:43:41.699
that were using that adapter.

00:43:41.699 --> 00:43:43.699
So the switch says, oh, because think about it,

00:43:43.699 --> 00:43:46.699
the adapter that failed was on interface one.

00:43:46.699 --> 00:43:49.699
The one that took its place is on interface five.

00:43:49.699 --> 00:43:53.699
You don't want the switch to keep sending traffic to interface one or to

00:43:53.699 --> 00:43:55.699
say, oh, interface one went down.

00:43:55.699 --> 00:43:57.699
I don't know where those MACs are anymore.

00:43:57.699 --> 00:44:00.699
You put interface, you know, two in its place,

00:44:00.699 --> 00:44:02.699
attach it to interface five on the switch,

00:44:02.699 --> 00:44:08.699
and then the host is going to send out packets using the MAC address for

00:44:08.699 --> 00:44:12.699
the source MAC address for the VMs, one for, you know, all the VMs.

00:44:12.699 --> 00:44:13.699
So if you have 30 VMs on there,

00:44:13.699 --> 00:44:17.699
it sends out 30 packets for those different VMs to interface five on that

00:44:17.699 --> 00:44:18.699
switch.

00:44:18.699 --> 00:44:23.699
Now the switch just learned where those VMs are and how to get back to

00:44:23.699 --> 00:44:24.699
that.

00:44:24.699 --> 00:44:26.699
So you probably want to notify your switches most of the time.

00:44:26.699 --> 00:44:28.699
It's not very often there will be a reason not,

00:44:28.699 --> 00:44:29.699
so you don't want to do that.

00:44:29.699 --> 00:44:31.699
Failed back.

00:44:31.699 --> 00:44:36.699
So my NIC one goes down, five takes over.

00:44:36.699 --> 00:44:39.699
And then my interface two comes back up again.

00:44:39.699 --> 00:44:40.699
What do I do?

00:44:40.699 --> 00:44:45.699
Do I fail back, meaning put five back in my standby and two back in my

00:44:45.699 --> 00:44:48.699
active, or does two become my new standby?

00:44:48.699 --> 00:44:51.699
That's kind of up to you.

00:44:51.699 --> 00:44:52.699
People sometimes will say, yes,

00:44:52.699 --> 00:44:56.699
fail back because they want to have a predictive setup.

00:44:56.699 --> 00:45:00.699
They want to know one and two are always my active and three and four are

00:45:00.699 --> 00:45:01.699
always my standby.

00:45:01.699 --> 00:45:04.699
And so if two goes down, it fails over to three.

00:45:05.699 --> 00:45:08.699
But when two comes back, because I got an alert and I fixed the problem,

00:45:08.699 --> 00:45:10.699
I know one and two are active again.

00:45:10.699 --> 00:45:14.699
So if I want to know a predictive setup for all my hosts,

00:45:14.699 --> 00:45:16.699
you might want to fail back.

00:45:16.699 --> 00:45:18.699
And if you fail back, what's going to do,

00:45:18.699 --> 00:45:22.699
it's going to send out packets again from those VMs on that adapter to,

00:45:22.699 --> 00:45:25.699
to whatever interface on the switch is plugged into the switch changes,

00:45:25.699 --> 00:45:28.699
it's Mac address table and start sending traffic over there.

00:45:28.699 --> 00:45:32.699
You may not want to fail back because that predictive,

00:45:33.699 --> 00:45:37.699
knowing the configuration isn't as important to you as just a stable

00:45:37.699 --> 00:45:38.699
environment.

00:45:38.699 --> 00:45:41.699
I don't want it to be flapping back and forth and changing back and forth.

00:45:41.699 --> 00:45:44.699
Just if it's working, leave it alone.

00:45:44.699 --> 00:45:46.699
I'm fine with two being my standby.

00:45:46.699 --> 00:45:51.699
I know that my active adapters are going to be one and two or three and

00:45:51.699 --> 00:45:52.699
four.

00:45:52.699 --> 00:45:53.699
It's going to be one of them.

00:45:53.699 --> 00:45:55.699
I'm fine with it being whichever ones it is,

00:45:55.699 --> 00:45:57.699
as long as I know it's one of these four, that kind of thing.

00:45:57.699 --> 00:45:59.699
Fail over order.

00:45:59.699 --> 00:46:03.699
You can set, as we discussed in our last demo, which ones are my active,

00:46:03.699 --> 00:46:07.699
which ones are standby and which ones are not used at all.

00:46:07.699 --> 00:46:10.699
I can set adapters to also just not be used on a port group.

00:46:10.699 --> 00:46:13.699
They're on the switch, but they're not on the port group.

00:46:13.699 --> 00:46:17.699
So NetFlow,

00:46:17.699 --> 00:46:21.699
NetFlow is we can configure NetFlow on a distributed virtual switch.

00:46:21.699 --> 00:46:26.699
And basically what it does is it sends layer three traffic from a switch

00:46:26.699 --> 00:46:28.699
to a NetFlow appliance.

00:46:28.699 --> 00:46:32.699
So setting up a NetFlow appliance is a whole separate outside of,

00:46:32.699 --> 00:46:34.699
you know, a vSphere configuration.

00:46:34.699 --> 00:46:37.699
It could be a VM, but it's not a vSphere configuration.

00:46:37.699 --> 00:46:39.699
You set up a NetFlow server,

00:46:39.699 --> 00:46:42.699
and then you can configure this port group to use that particular NetFlow

00:46:42.699 --> 00:46:43.699
server.

00:46:43.699 --> 00:46:45.699
And it sends information,

00:46:45.699 --> 00:46:48.699
samplings to that NetFlow server to give you information about your

00:46:48.699 --> 00:46:50.699
network, okay?

00:46:50.699 --> 00:46:51.699
And block ports.

00:46:51.699 --> 00:46:53.699
You can say you want to block all the ports or block none.

00:46:53.699 --> 00:46:56.699
So you might just want to block, you know,

00:46:56.699 --> 00:46:59.699
bring down that virtual switch for a moment without actually unplugging

00:46:59.699 --> 00:47:00.699
things or anything like that.

00:47:00.699 --> 00:47:01.699
You can block all the ports.

00:47:01.699 --> 00:47:03.699
And then when you're done doing your maintenance,

00:47:03.699 --> 00:47:06.699
you can unblock them.

00:47:06.699 --> 00:47:08.699
So verify.

00:47:08.699 --> 00:47:10.699
Once you have, you know, done a configuration,

00:47:10.699 --> 00:47:12.699
if you go to your host, like, well, it's not there.

00:47:12.699 --> 00:47:13.699
I don't see it.

00:47:13.699 --> 00:47:14.699
Go to your tasks.

00:47:14.699 --> 00:47:18.699
You'll see at the bottom of your vSphere client that when you say okay

00:47:18.699 --> 00:47:20.699
to something, it creates a task.

00:47:20.699 --> 00:47:22.699
It shows you the progress of that task.

00:47:22.699 --> 00:47:24.699
Maybe it hasn't completed yet.

00:47:24.699 --> 00:47:28.699
Deploying a new virtual switch to hosts can be somewhat time consuming,

00:47:28.699 --> 00:47:30.699
especially if you did quite a few.

00:47:30.699 --> 00:47:32.699
And so it could take a little bit.

00:47:32.699 --> 00:47:35.699
So just watch the task, wait for it to complete completely.

00:47:35.699 --> 00:47:37.699
Then you can go to the host and refresh,

00:47:37.699 --> 00:47:39.699
and you should see all the changes there, okay?

00:47:39.699 --> 00:47:41.699
Check the ports on your switch.

00:47:41.699 --> 00:47:43.699
If you are using distributed virtual switches,

00:47:43.699 --> 00:47:46.699
and if you configured things like lags and things like that,

00:47:46.699 --> 00:47:49.699
you need to configure your switch to also use those lags.

00:47:49.699 --> 00:47:53.699
A standard switch cannot do a dynamic lag.

00:47:53.699 --> 00:47:55.699
A standard switch is going to do what's called a static lag.

00:47:55.699 --> 00:47:59.699
A distributed virtual switch can do either a static lag

00:47:59.699 --> 00:48:01.699
or an LACP dynamic lag.

00:48:01.699 --> 00:48:04.699
So however you have your virtual switches configured,

00:48:04.699 --> 00:48:09.699
make sure that your ports on your physical switch match that configuration.

00:48:09.699 --> 00:48:10.699
Same thing with VLANs.

00:48:10.699 --> 00:48:13.699
Make sure all your VLANs that you configured on your port groups

00:48:13.699 --> 00:48:19.699
and things are tagged on the switch itself, the physical switch.

00:48:19.699 --> 00:48:22.699
So in conclusion, we discussed standard virtual switches,

00:48:22.699 --> 00:48:24.699
and we discussed distributed virtual switches.

00:48:24.699 --> 00:48:26.699
Before we wrap up this lesson,

00:48:26.699 --> 00:48:28.699
let's go ahead and go back to our demo environment,

00:48:28.699 --> 00:48:33.699
and I'll show you how to set up and configure a distributed virtual switch.

00:48:33.699 --> 00:48:35.699
So here we are back in our demo environment.

00:48:35.699 --> 00:48:38.699
We have our Houston data center.

00:48:38.699 --> 00:48:42.699
We have our production cluster here, and I've got a couple of hosts in here,

00:48:42.699 --> 00:48:51.699
and these hosts have some unused unclaimed adapters on both these hosts.

00:48:51.699 --> 00:48:54.699
One thing that's really cool, I can point out here too,

00:48:54.699 --> 00:48:56.699
switch nothing claimed here also,

00:48:56.699 --> 00:48:59.699
is notice that, like, for example, when I was here

00:48:59.699 --> 00:49:04.699
and I went to adapters and showed you, when I click on the other host,

00:49:04.699 --> 00:49:05.699
I stayed on that same screen.

00:49:05.699 --> 00:49:09.699
It didn't take me, for example, to the summary screen right away, right?

00:49:09.699 --> 00:49:14.699
So if I go to this host now, notice when I stayed on my summary screen,

00:49:14.699 --> 00:49:18.699
I didn't jump back to where I was last in this host.

00:49:18.699 --> 00:49:19.699
It took me to wherever I was.

00:49:19.699 --> 00:49:22.699
It makes it really easy if I'm trying to configure certain things on multiple

00:49:22.699 --> 00:49:25.699
hosts to say I want to look at virtual switches.

00:49:25.699 --> 00:49:26.699
Okay, I see this.

00:49:26.699 --> 00:49:29.699
Let me compare it to this guy here and take me to the same spot on that host.

00:49:29.699 --> 00:49:32.699
That's a really cool feature that I really like.

00:49:32.699 --> 00:49:35.699
So from here I'm going to go to my networking option in vSphere.

00:49:35.699 --> 00:49:39.699
I have my Houston data center, right,

00:49:39.699 --> 00:49:41.699
and I can see the networks that are configured on there.

00:49:41.699 --> 00:49:46.699
So here I want to say I want to create a distributed switch,

00:49:46.699 --> 00:49:49.699
and I want to create a new distributed virtual switch.

00:49:49.699 --> 00:49:52.699
I'm going to give it a name.

00:49:52.699 --> 00:49:56.699
I'm going to call this distributed virtual switch,

00:49:56.699 --> 00:49:59.699
and I'm going to put it in that data center.

00:49:59.699 --> 00:50:02.699
Now, this is where you choose the version.

00:50:02.699 --> 00:50:06.699
I know all my hosts are going to be eight, so I can say eight or later.

00:50:06.699 --> 00:50:11.699
If I knew I had some seven update three hosts still in my environment

00:50:11.699 --> 00:50:14.699
or later, then I might choose that option, right?

00:50:14.699 --> 00:50:15.699
So I'm going to go ahead and choose eight.

00:50:15.699 --> 00:50:18.699
I got all version eight hosts in this environment.

00:50:18.699 --> 00:50:20.699
I'm going to say next.

00:50:20.699 --> 00:50:24.699
So are we doing the offload capabilities?

00:50:24.699 --> 00:50:27.699
So what that basically means is do we have a smart NICs?

00:50:27.699 --> 00:50:28.699
Do we have, like, you know,

00:50:28.699 --> 00:50:33.699
NICs that are able to do ISCSI offloading or TCP offloading or, you know,

00:50:33.699 --> 00:50:36.699
SSL offloading, that kind of thing.

00:50:36.699 --> 00:50:37.699
I'm going to say none.

00:50:37.699 --> 00:50:39.699
Number of uplink ports, I'll do four.

00:50:39.699 --> 00:50:42.699
Network IDLE control, you can enable it or not.

00:50:42.699 --> 00:50:43.699
That's an option for you

00:50:43.699 --> 00:50:45.699
if you're going to be using the network IDLE control or not.

00:50:45.699 --> 00:50:48.699
If you're not, you can go ahead and turn it off.

00:50:48.699 --> 00:50:50.699
Do you want to create that default port group?

00:50:50.699 --> 00:50:52.699
I'm going to go ahead and uncheck that and just go ahead

00:50:52.699 --> 00:50:56.699
and we'll create one after we're done.

00:50:56.699 --> 00:50:57.699
And I'll say next.

00:50:57.699 --> 00:50:58.699
And it's all set.

00:50:58.699 --> 00:51:03.699
It's going to go ahead and create my distributed virtual switch.

00:51:03.699 --> 00:51:06.699
And that happened actually really fast.

00:51:07.699 --> 00:51:14.699
So from here, I have that DemoDB uplink port group there, our uplink group.

00:51:14.699 --> 00:51:20.699
I'm going to go in and say I want to create a distributed port group, right,

00:51:20.699 --> 00:51:22.699
a new distributed port group.

00:51:22.699 --> 00:51:24.699
So this will be my port group,

00:51:24.699 --> 00:51:30.699
and I'll just call it a distributed port group demo.

00:51:30.699 --> 00:51:36.699
I'm going to put it on that demo distributed virtual switch.

00:51:36.699 --> 00:51:39.699
Now, here are all the settings we talked about in our lesson.

00:51:39.699 --> 00:51:43.699
All right, do we want to do port bindings, right, ephemeral, no binding?

00:51:43.699 --> 00:51:44.699
That's usually fine.

00:51:44.699 --> 00:51:46.699
You can say static if you wanted to, right.

00:51:46.699 --> 00:51:51.699
If you're going to say you want it to be bound, elastic, what size,

00:51:51.699 --> 00:51:53.699
notice they give you a really small number to start.

00:51:53.699 --> 00:51:57.699
If you're just going to say it's going to be a fixed size, then it says,

00:51:57.699 --> 00:51:59.699
well, then you might want to choose a lot of ports.

00:51:59.699 --> 00:52:01.699
But if you're going to say it's elastic so it can grow,

00:52:01.699 --> 00:52:03.699
you start with a small number of ports,

00:52:03.699 --> 00:52:06.699
and that's the increment in which it's going to grow in, okay.

00:52:06.699 --> 00:52:09.699
Are we going to do a network resource pool?

00:52:09.699 --> 00:52:11.699
I'm not going to assign a network resource pool.

00:52:11.699 --> 00:52:14.699
I don't want to, you know, limit the amount of IO it's going to use

00:52:14.699 --> 00:52:16.699
and things like that.

00:52:16.699 --> 00:52:21.699
VLAN, now here you might choose a VLAN that you've configured before

00:52:21.699 --> 00:52:24.699
in a distributed switch, a standard distributed switch,

00:52:24.699 --> 00:52:29.699
and now we're going to migrate this to this distributed virtual switch

00:52:29.699 --> 00:52:31.699
from a standard switch, right.

00:52:31.699 --> 00:52:34.699
Do you want to customize any of these other policies that we talked about?

00:52:34.699 --> 00:52:37.699
Do you want to do, you know, security settings

00:52:37.699 --> 00:52:40.699
and, you know, MAC address spoofing, that kind of thing?

00:52:40.699 --> 00:52:44.699
Right now we just go ahead and say next and finish.

00:52:44.699 --> 00:52:49.699
And there is my port group.

00:52:49.699 --> 00:52:51.699
I can say edit settings here.

00:52:51.699 --> 00:52:56.699
I can look at my teaming and failover to have my uplinks,

00:52:56.699 --> 00:52:59.699
and here it's showing that all my uplinks on this group are active.

00:52:59.699 --> 00:53:04.699
I can choose, you know, three and four, move those down, they're standby.

00:53:04.699 --> 00:53:10.699
I can create another port group and make three and four my active

00:53:10.699 --> 00:53:13.699
and one and two standby for that one, just like we did on the standard switch.

00:53:13.699 --> 00:53:16.699
I'll say next.

00:53:16.699 --> 00:53:18.699
It's okay there.

00:53:18.699 --> 00:53:22.699
Now I reconfigured that, and that's completed successfully.

00:53:22.699 --> 00:53:28.699
So now what I can do, I can add VM kernel adapters to here if I wanted to.

00:53:28.699 --> 00:53:33.699
I can migrate VMs to this port group that are on, you know,

00:53:33.699 --> 00:53:37.699
other configurations, other standard switches.

00:53:37.699 --> 00:53:42.699
On the standard switch itself, I can do that.

00:53:42.699 --> 00:53:49.699
On the standard switch itself, I can say add managed hosts.

00:53:49.699 --> 00:53:50.699
So I'm going to add hosts.

00:53:50.699 --> 00:53:52.699
Managed hosts need to be changing settings for hosts that are already here.

00:53:52.699 --> 00:53:56.699
I have no hosts here yet, so I'm just going to add some hosts.

00:53:56.699 --> 00:54:01.699
I'm going to add these two hosts.

00:54:01.699 --> 00:54:03.699
Which adapters do I want to use?

00:54:03.699 --> 00:54:10.699
I'm going to say I want four to go to uplink one, five to go to uplink two,

00:54:10.699 --> 00:54:15.699
six to go to uplink three, and seven to go to uplink four, right?

00:54:15.699 --> 00:54:17.699
And that's because my adapter started at zero.

00:54:17.699 --> 00:54:21.699
So I have eight on here, and zero is the first one,

00:54:21.699 --> 00:54:22.699
one is the second one, and so forth.

00:54:22.699 --> 00:54:25.699
So four is actually five, six, seven, and eight.

00:54:25.699 --> 00:54:28.699
So uplinks one, two, three, and four.

00:54:28.699 --> 00:54:30.699
I'm going to say next.

00:54:30.699 --> 00:54:32.699
And are we going to move any VM kernel ports?

00:54:32.699 --> 00:54:34.699
I don't have to move any of these, right?

00:54:34.699 --> 00:54:38.699
I can leave them on the standard switches they're on, and that will work fine.

00:54:38.699 --> 00:54:43.699
If I wanted to create a VM kernel port on my distributed virtual switch

00:54:43.699 --> 00:54:45.699
and assign it somewhere, I could do that.

00:54:45.699 --> 00:54:49.699
If I wanted to migrate virtual machines that were on the standard switch to

00:54:49.699 --> 00:54:55.699
over to my kernel, my port groups on this switch, I could do that as well.

00:54:55.699 --> 00:54:58.699
But we're not going to migrate any of the VMs.

00:54:58.699 --> 00:55:02.699
And we can say finish.

00:55:02.699 --> 00:55:05.699
So we're reconfiguring the vSphere distributed switch, right?

00:55:05.699 --> 00:55:09.699
We're adding the host, and that completed.

00:55:09.699 --> 00:55:14.699
So if I go back to my host, and if I were to look at virtual switches,

00:55:14.699 --> 00:55:18.699
now you see I have the distributed virtual switch here, right?

00:55:18.699 --> 00:55:20.699
I have the distributed virtual switch.

00:55:20.699 --> 00:55:27.699
I have my uplinks on there, all configured, right?

00:55:27.699 --> 00:55:29.699
So that's pretty cool, right?

00:55:29.699 --> 00:55:34.699
I was able to deploy a switch to a host, right?

00:55:34.699 --> 00:55:35.699
Same thing here.

00:55:35.699 --> 00:55:37.699
I deployed this switch to this host.

00:55:37.699 --> 00:55:40.699
It's got the uplinks all configured on there, and it's all ready to go.

00:55:40.699 --> 00:55:41.699
So that's pretty exciting.