1.2 Virtual Machines: Guests of the Hypervisor

00:00

Hi,

00:02

you're watching VM one of these fear lesson to virtual machines.

00:06

By the end of this video, you'll learn what virtual machine is and how it gets access to resource is in order to run.

00:14

Before we talk about virtual machines, let's quickly review what the high advisor is.

00:19

The hyper visor is an operating system like any other operating system,

00:24

in the sense that it is installed on a bare metal server. And just like any other operating system, it's monopolizes the hardware.

00:32

But unlike any other operating system, it hosts other operating systems,

00:37

and then it tricks them into believing they are running directly on the hardware.

00:42

It's a cunning trickster.

00:44

The hyper visors ultimate drop they stand to share. The physical resource is among the virtual machines.

00:51

Now the virtual machine ISS. Like any regular machine, it's made often operating system and one or more applications. But it is unlike any other regular machine in the sense that it is a guest off. Another operating system,

01:06

which is the hyper visor

01:07

virtual machine, has no monopoly over the hardware. In fact, it shares the underlying hardware with other visual machines.

01:15

So the PM's Traube well,

01:19

it's to eat up Resource is served by its host

01:22

virtual machine needs. CPU memory storage and networking, among other resource is for its normal operation. But where does it get those? Resource is from

01:34

well from the host, of course.

01:36

But being a guest off another operating system, a virtual machine is oblivious to the amount of resources available in the physicals ever. So whatever the hyper visor presents to the V M, the VM believes it's all varies.

01:51

The V M always gets These resource is through the heart of Russia We call these virtual resource is as opposed to physical resources,

02:00

and here's where it gets interesting. The hyper visor itself will install vendor specific drivers in order to tap into the real hardware. The physical hardware,

02:09

if then presents generic drivers to the guest operating system in order for the VM to use the virtual hardware.

02:19

This makes the VM completely decoupled from the physical hardware underneath, and this will prove extremely useful for portability.

02:27

Now the hyper visor also provides additional virtual hardware to the PM's, including a CD or DVD drive, a USB controller, a scuzzy or satar disk controller, a keyboard, a mouse et cetera.

02:42

Now, except for the CPU, the rest off a virtual machines hardware is actually emulated.

02:49

What you see in the list above our emulation sze, created by the hyper visor.

02:53

But remember, when it comes to the CPU off any virtual machine that is no emulation taking place,

03:01

the PM's are being scheduled to run on the physical CPU.

03:06

Let's not delve into complicated technicalities, but this was important. To point out, let's now see an example of how the heart provides the shares. Resource is among the EMS.

03:16

Here's our physical server configuration. It's made up of two quad core CPI use, 32 gigabytes of memory, one terabyte of storage space and one network interface card. These resources are monopolized by the hyper visor.

03:31

Now let's say we have three virtual machines. VM one gets to virtual CB. Use four gigs of RAM, 200 kicks off this and one virtual nick VM two gets one virtual CPU. Two gigs of RAM, 60 forgets of disk and one V. Nick Vien three

03:50

gets eight. Virgil CB used six kicks off Ram,

03:53

500 kicks of disk and to Phoenix. Let's talk about CPU for a moment

03:58

in fact, when it comes to CPU allegation, one virtual CPU equals one logical CPU. Now, what's a logical CBU? Well, most CP use can leverage something called hyper threading, which is the ability to execute two routines or threats simultaneously.

04:16

So a single physical CPU core with hyper threading enabled appears as too logical. C p usedto an operating system.

04:26

So our eight cores, with two threats, each become 16 Logical Seabees.

04:31

No,

04:32

let me ask you a question.

04:34

What's the maximum number of easy P use Any visual gene can get?

04:40

Well,

04:41

it's the total number off logical CP use in the server.

04:44

You can't exceed that.

04:46

But there is also a hard limit specified by the vendor, which, in the case of E. M. Where is 128

04:54

now, let's go back to our diagram. Given that we have 16 logical CB use, each virtual machine can be configured with up to,

05:01

well, 16 virtual see piers.

05:05

Now, let me ask you another question.

05:08

What's the maximum amount of memory for Vian?

05:11

Well, it's the total amount of physical memory

05:15

minus the overhead memory used by the hyper visor.

05:18

Naturally, the hyper visor needs memory to operate itself.

05:23

In fact, we'll also have to sub strapped the memory overhead used by the hyper visor to run all the V ins.

05:30

But let's ignore that for now

05:32

again. There is also a hard limit for memory, which is in the case off camera six terabytes Peruvian

05:41

Back to our diagram. If we suppose the hyper visor overhand is two gigs and the total memory off the server is 32 games, then we're left with 30 gigs of RAM,

05:51

so every virtual machine can be configured with upto 30 kicks off ram. As long as you do not configure any reservations,

06:00

we'll talk about reservations in another module.

06:03

But

06:03

wait a minute.

06:05

The total amount of allocation off memory across all three v EMS is 30 plus 30 plus 30 which adds up to 90 gigabytes of RAM.

06:15

This is greater than the total memory in the server.

06:19

How is that possible?

06:20

Well, we call this memory over commitment,

06:25

and we'll talk about this in the resource management lesson.

06:29

The same thing goes for storage. If the hyper visor overhead is six gigs,

06:33

we're left with 994 gigs off space,

06:36

which is the amount we can configure for each of the sweet virtual machines, but again, without any reservations.

06:45

And if we do configure this on all V EMS, we will end up with a storage over commitment, because this some off allegations is greater than what to the server can supply in terms of storage space. Finally, regarding the maximum number of Phoenix pervy, um, the answer is going to be different.

07:02

A virtual machine is not limited by the number off physical next on the server,

07:08

because the hyper visor can emulate network interface cards and then present them into the V EMS.

07:14

So regardless of the number of physical necks, the maximum V necks you can configure on of'em in the case off the M, where is 10?

07:23

Let's have a final look at our diagram.

07:25

These are the configuration maximums pervy um, in our case,

07:30

one important thing to remember.

07:31

Do not play around with resource allocation because you could easily end up in an overcommitted environment that is very complicated to troubleshoot, especially when you have plenty of Ian's running. Always stick with whatever your application vendor recommends.

07:48

Well, this wraps up our introduction to virtual machines here are important takeaways.

07:55

Virtual Machine is a guest machine that crumbs on top of the hyper visor. It consumes virtual resources that are nothing more than a portion of the physical resources. Virtual machine is also extremely easy to migrate and move around because it is hardware independent.

08:11

Thanks to the generic drivers provided by the hyper visor,

08:16

the hyper visor is what hosts the PM's

08:18

and its main job is to share. The physical resource is among the virtual machines.