1
00:00:00,740 --> 00:00:04,760
Lastly, we have Hyper‑V network load balancing, and the idea here,

2
00:00:04,760 --> 00:00:08,020
this is a Windows Server service where we're implementing load

3
00:00:08,020 --> 00:00:12,380
balancing in software in the operating system where we have two or

4
00:00:12,380 --> 00:00:16,430
more identically configured Windows Servers with the NLB role

5
00:00:16,430 --> 00:00:19,160
installed, and you form a cluster there. Again,

6
00:00:19,160 --> 00:00:24,340
your advertising connectivity to that server farm on a virtual IP address

7
00:00:24,340 --> 00:00:29,950
and NLB just simply allows traffic coming in on that virtual IP to be round

8
00:00:29,950 --> 00:00:33,150
robin‑ed from one backend node to the next.

9
00:00:33,150 --> 00:00:37,750
It's not a particularly robust or intelligent type of load balancing,

10
00:00:37,750 --> 00:00:41,160
but it's been part of Windows Server for a long time, and you can't

11
00:00:41,160 --> 00:00:43,600
beat the price. If you're familiar with Azure,

12
00:00:43,600 --> 00:00:46,880
there are also plenty of software‑defined networking

13
00:00:46,880 --> 00:00:49,080
load balancing options there as well.

14
00:00:49,080 --> 00:00:52,810
They all have the same idea though. If one or even two of

15
00:00:52,810 --> 00:00:55,730
these web VM machines were to go offline,

16
00:00:55,730 --> 00:01:06,000
as long as we had at least one healthy node left in the server farm, the service would still be available and accessible on that virtual IP address.

