2 hours 28 minutes

Video Description

This module teaches about virtual machine (VM) and host resource allocation:

  1. Resources: memory, CPU and disk space
  2. Quotas: hard and soft limits
  3. Licensing
  4. Reservations
  5. Resource Pools
  6. Dynamic Resource Allocation
  7. Physical Resource Redirection

The module also covers: 1. Common Disk Failures

  1. Common Memory failures
  2. Network failures

The module also teaches about common performance considerations: 1. Disk

  1. Caching
  2. Network
  3. Bandwidth
  4. Latency
  5. Quality of Service (QoS)
  6. Multi pathing
  7. Load balancing

Video Transcription

Welcome back to Cyberia. IittIe My name is Justin Lingam. Today we'll be talking about allocation and optimizing
is an overview. We're gonna talk about the VM and host resource allocation configuration practices, common issues and common performance concepts with the human host resource allocation won't talk about the resource is quota licensing, reservation resource pool, dynamic resource allocation and physical resource. Redirect
on resource is common. Resource is that we have our memories, CPU and dis space. These resources allow the VM to be able to perform that you have to have physical round. You have to have physical CPU for the V M.
And if you don't have enough CP around, then you're not going to be able to host another Veum. At the same time. Your dis space you need to have enough display. So that way you can give it to a VM. If you're loading Windows seven V M. Which uses about 20 to 30 gigs, then you have to have 20 to 30 gigs available. For that Veum to take
at the same time. It needs at least one core
and one giga memory. But if that Liam is going to be running programs on it,
then it needs more than one core, and it needs more than one gig of memory, so you need to have that available.
Most servers that hosted the EMS, however, have around
40 course at the same time, they have up to 120 gigs of RAM space.
Next day we don't talk about its quota quota is basically a limit on how many resource is the Veum can use. There are hard and soft limits when it comes to quota. The soft women is a limit set but can't be exceeded. For example, your Windows seven of'em that you're running, you said a soft women of two gigs of RAM for they've am to use.
But you said a hard limit of four gigs.
Well, as the VM starts using its resource is, it might hit that two gates and then exceed that, which is okay. But as soon as it is that four gigs, that's where it stopped, and it's not able to go any further. Licensing is usually the most expensive part. Whenever it comes to any organization.
For example, Microsoft Word, you can actually get a free trial of Microsoft, where for so many days and then you can use it. And then once that free trial ends,
you connect your limited on the features you can use. But she was still use Mark software. However, the license can cost up to $120 just for that license, so the license is far exceeds the actual cost of the project. That's true for anything you can download Window seven for free,
but the license for it could be 70 $80 just for Windows seven Veum, where you can get Veum wear for free
and you'll be able to use it last you for the license key.
But then the license. Kiefer Veum. Where could be in the thousands whenever it comes to an organization? Reservations reservations are a set aside of resource is that the customer is given to a particular VM.
Reservations are usually stated within the contract of the company. The company's contract says that they will get three machines Well, then they will get those three machines.
The same goes for the actual CPU uses or the number of CPU is that they get
resource pool Resource pool is very sandwich. Reservations, however, is a pool of resource is that the company's Everly used. If their contract states that they get 20 CPU course will, then in this case, they're going to get all 20 city. Of course, that is a reservation, and that is also there, pull that they're going to receive.
So if they need all 20 city, of course,
well, then they will get all 20 CP. Of course, the purpose behind Resource Pool it's provide the resource is organized for the company and the provider
Nice, and we will talk about his dynamic resource allocation.
Dynamic means changing. So dynamic resource allocation is whenever you provide a computer with the ability to change based on what it's needed. If the computer is using more resource is, it will dynamically change it so that away it can give it more resource is that it needs. The resource is might be CPU, memory displaced
or other things
as, for example, the nick. The next time we talk about its physical resource, redirect physical resource. Redirect is whenever you have an actual poor on a computer, the host say USB port and your redirected to the V M. The reason why you wanna have that is maybe for the keyboard.
You want the venom to use the keyboard, So therefore, you d o a
resource redirect at the same time. Maybe you have an external hard drive when you want the VM to utilize that external hard drive. So you do a resource Redirect
resource. Redirect isn't limited. Just the U. S. Fees. You can also use the parallel port for printers. Or you can use a serial port for other devices, like maybe a trackball, maybe whatever computer, using like a cat,
needs a trackball and configuration practice. We're gonna talk about memory ballooning. We won't talk about the CPU swap this space and dis performance Memory is usually the limiting factor whenever it comes to a virtual machine. Memory blooming allows you to exceed that. The operating system uses third party software
to find out if it needs that have more memory or not.
Memory blooming gives you a little bit more memory so that it were your ability to utilize everything to its full extent swap. This space is very similar to page fall, the swat. This space
is whenever the Veum is using all of its memory and then it needs to use a little bit extra, but there is no memory available to it, so therefore, it uses a little bit of the hard drive space
for memory.
It does call some performance issues, but it does allow you to use more memory
essentially on that on that V M
CPU sippy use have course in them, and each the needs a particular core. If you have a quad core machine and you're using your trying to run for V EMS, you're going to have
you're gonna overtax that CPU
and you're gonna have performance issues. If you have a quad core, you can only run three v EMS. One corps is gonna be utilized just to the host so it can perform and be able to do the normal functions that it needs to do at the same time. The faster the CPU, the more you're able to perform with these v EMS.
When it comes to Williams, you do need you dis space. And if you're running out of this space you're running low on dis space or you're accessing your hard drive to too much. Then you could be causing issues inside of your VMS because performance issues.
On the other hand, raid allows you to really V eight that, if you remember there a few different types of rate rate. Five is the most common one, where it's able to utilize three separate hard drives and access information quicker on three hard drives versus one hard drive. Next, we're gonna talk about some common failures with the disk CPU and the network
common desk failures. We have the physical disc
failure. You also have descriptions and h b A failures. Physical hard. This failure could be a sector on a hard drive that is going bad. And then, whenever information is being written to that, it gets corrupted. And then it's not able thio to utilize that sector anymore. One way to relieve E eight that it's around a cheque disk on the physical hard disk,
and what this allows you to do is
check the each sector to see if there are any heirs, and if there are errors, then it is noted in the
and the concern and the hard drive's controller.
So that way it knows not to write to that person anymore.
The H B A would be similar to the actual poor that your motherboard is using for the hard drive
and a failure with that. What actually caused that port to be unusable? Some common 1,000,000 memory failures are the ram failures and swap fallout of space. With the round, the chips do go bad and the modules do need to be replaced. However, to replace those modules is the hard process. So replacing the round
is more efficient
for the network. You do have common failures in it. With the speed and the duplex, the speed will be. If you're networkers running, get 100 mags. But you have your Veum running at 10 makes you're going to cause a network failure. Or the duplex. The duplex. You have single, half and full do place
if your network is running at half to place. But you have it running at full duplex, you're going to cause
some more failures. And then the physical Nick, you could have failures with it. And if the physical Nick fails, well, then you're gonna have failures in all of the EMS.
So if all volumes are not responding correctly, then you might want to check the actual host and see if it's having issues.
Lastly, we have some considerations whenever it comes to performance. The disc cash Network band with Leighton see quality of service, multi passing and load balancing. Each one offers you the ability to increase your networks performance. You increase your dis size or you increase your dis speed than you're able to write more.
Cashing allows your computer to do more local operations
and not have to go to the Internet.
The faster your network, the more you're able to give. Resource is out to all users.
Bandwidth allows you to give more. Resource is out as a whole. Reducing Listen, See allows your customers to see an improved performance. The lower the late and see means that the faster they're getting their information.
Quality of service is whenever you provide voice or other VP ends with quality and allows them to be
toe have the highest priority.
You want to give your voice voice over I p
the highest priority. So that way they're able to communicate and do what they need to dio. Multi patching gives you redundancy whenever it comes to the storage devices and allows you to have many routes. That storage device, though balancing allows youto have many ways of getting information out. For example, to routers versus one router
are to fall servers that are replicating back and forth versus one
fall server. And this module we covered vehement hosts, resource allocation configuration, best practices, common issues and common performance concepts. When it was just Lingle with library, I t know if you guys learned a lot.

Up Next

CompTIA Cloud+

Our online Cloud+ training addresses the essential knowledge for implementing, managing and maintaining cloud technologies as securely as possible. It covers cloud concepts and models, virtualization, and infrastructure in the cloud.

Instructed By

Instructor Profile Image
Justin Langham
System and Network Engineer