REVISE: Hyper-V Resource Control and Resource Metering

Very impressive links on Hyper-V Resource Control and Resource Metering.

Question: Windows Server 2012 and 2012 R2

IT organizations need tools to charge back business units that they support while providing the business units with the right amount of resources to match their needs. For hosting providers, it is equally important to issue chargebacks based on the amount of usage by each customer.

To implement advanced billing strategies that measure both the assigned capacity of a resource and its actual usage, earlier versions of Hyper-V required users to develop their own chargeback solutions that polled and aggregated performance counters. These solutions could be expensive to develop and sometimes led to loss of historical data.

To assist with more accurate, streamlined chargebacks while protecting historical information, Hyper-V in Windows Server 2012 introduces Resource Metering, a feature that allows customers to create cost-effective, usage-based billing solutions. With this feature, service providers can choose the best billing strategy for their business model, and independent software vendors can develop more reliable, end-to-end chargeback solutions on top of Hyper-V.

Key benefits

Hyper-V Resource Metering in Windows Server 2012 allows organizations to avoid the expense and complexity associated with building in-house metering solutions to track usage within specific business units. It enables hosting providers to quickly and cost-efficiently create a more advanced, reliable, usage-based billing solution that adjusts to the provider’s business model and strategy.

Use of network metering port ACLs

Enterprises pay for the Internet traffic in and out of their data centers, but not for the network traffic within their data centers. For this reason, providers generally consider Internet and intranet traffic separately for the purposes of billing. To differentiate between Internet and intranet traffic, providers can measure incoming and outgoing network traffic for any IP address range, by using network metering port ACLs.

Virtual machine metrics

Windows Server 2012 provides two options for administrators to obtain historical data on a client’s use of virtual machine resources: Hyper-V cmdlets in Windows PowerShell and the new APIs in the Virtualization WMI provider. These tools expose the metrics for the following resources used by a virtual machine during a specific period of time:

  • Average CPU usage, measured in megahertz over a period of time.
  • Average physical memory usage, measured in megabytes.
  • Minimum memory usage (lowest amount of physical memory).
  • Maximum memory usage (highest amount of physical memory).
  • Maximum amount of disk space allocated to a virtual machine.
  • Total incoming network traffic, measured in megabytes, for a virtual network adapter.
  • Total outgoing network traffic, measured in megabytes, for a virtual network adapter.

Movement of virtual machines between Hyper-V hosts—for example, through live, offline, or storage migrations—does not affect the collected data.

Question: Introduction to Resource Metering

Hi, I’m Lalithra Fernando, a program manager on the Hyper-V team, working in various areas including clustering and authorization, as well as with our Hyper-V MVPs. In this post, I’ll be talking about resource metering, a new feature in Hyper-V in Windows Server 2012.

As you’ve probably heard by now, Windows Server 2012 is a great platform for the private cloud. When we began planning this release, we realized that one of the things you need in order to run a cloud is to be able to charge your users for the resources they use.

This is the need resource metering fills. It allows you to measure the resource utilization of your virtual machines. You can use this information as a platform for your own dynamic chargeback solutions, where you can charge customers based on the resources they use instead of a flat upfront cost, or to plan your hosting capacity appropriately.

There are four resources that you can measure: your CPU, memory, network, and storage utilization. We measure these resources over the period of time between when you measure and when you last reset metering.

CPU (MHz): We report the average utilization in megahertz.

Now, you’re probably wondering why we don’t report this as a percentage. After all, that’s what we do in Hyper-V Manager. Well, we know that you like to move your virtual machines. With Windows Server 2012, you can live migrate your virtual machines all over the place. Naturally, the record of how much resources your virtual machine has used moves with it.

We want the virtual machine’s CPU utilization to make sense across multiple machines. If we report a percentage and you move the virtual machine to a host with different processing capabilities, it’s no longer clear what the percentage refers to.

Instead, we report in megahertz. For example, if your virtual machine had an average CPU utilization of 50% over the past billing cycle on a host with a CPU running at 3GHz, we would report 1500MHz.

If your virtual machine spent one hour on a host with a 3GHz CPU and used 50% and another hour on a host with 1GHz CPU and used 75%, we would report the following:

(3GHz * 1000MHz/1GHz * .5 * 1hr) + (1GHz * 1000MHz/1GHz * .75 * 1hr) = 2250MHz-Hr

Here I am converting the CPU capacity from GHz to MHz and figuring out how much of that capacity was used over each hour.

2250MHz-Hr / 2 Hours = 1125 MHz.

Then, I simply divide over the two hours to get this value.

One final note: we don’t report minimum and maximum utilization values for CPU. If you think on it a moment, you’ll come to the same realization we did: it is very likely that the minimum will be 0 and the maximum will be the full capacity of the hosts’ CPU at some point over the timespan you’re measuring. Since that’s not very useful, we don’t report it.

Memory (MB): We report the average, maximum, and minimum utilization here, in megabytes.

The minimum memory utilization captures the least memory used over the timespan measured. Since it’s not very useful to know that the minimum memory usage was zero if the virtual machine was ever turned off, we only look at the minimum memory utilization when the virtual machine is running.

We do include the offline time of the virtual machine when calculating the average memory utilization. This provides an accurate view of how much memory the virtual machine was using over that billing cycle, so that you can charge your users accurately.

Network (MB): We report network utilization in megabytes. Of course, we want this metric to be useful, so we considered how you would want to see this information broken down. One way you might want to distinguish between network traffic is to see how much traffic is inbound to the virtual machine, and how much is outbound.

The most important breakdown you will want is how much traffic does the virtual machine send or receive from the internet, which costs you money, and how much is just communication between virtual machines you host, which costs you nothing since it is just using your intranet. With this breakdown, you can charge your user accurately for their internet usage.

So how do we provide these breakdowns? We use ACLs set on the virtual machine’s network adapter. Each ACL has

  • Direction
    • “Inbound” or “Outbound”
  • Remote IP Address
    • The source or destination of the network packet, depending on direction
    • For example, 10.0.0.0/8
  • Action
    • Allow, Deny, or Meter

These ACLs are used for more than just resource metering; note the Allow and Deny actions. For our purposes, you set the action to “Meter”.

Enabling resource metering creates two sets of default metering ACLs, provided none are already configured. The first set of ACLs, one inbound and one outbound, has a remote IP address of *.*; this wildcard notation indicates that it will meter all IPv4 traffic that is not covered by another ACL. The second set has an IP address of *:*, which meters all IPv6 traffic.

With these metering ACLs, you can measure the total network traffic sent and received by the virtual machine, in megabytes. You can configure your own ACLs to count intranet traffic separately from internet traffic, and charge accordingly.

Disk (MB): As we spoke with customers, we realized that for chargeback purposes, they were only interested in the total disk allocation for a virtual machine. So, here we report that in megabytes.

The total value is the capacity (not the current size on disk) of the VHDs attached to the virtual machine plus the size of the snapshots. Take the following examples:

Fixed size disk: 

VM with a single 100GB fixed size VHDs attached

————————————————————-

Total Disk Allocation reported: 100GB

Dynamic disk:

VM with a single dynamic VHD attached, 

Current size 30GB, maximum size 100GB

————————————————————-

Total Disk Allocation reported: 100GB

With snapshots:

VM with a single dynamic size VHDs attached,

Current size 30GB, maximum size 100GB,

Plus a 20GB snapshot

————————————————————-

Total Disk Allocation reported: 120GB

Pass-through disks, DAS disks, guest iSCSI connections, and virtual Fibre Channel disks are not included in the total disk allocation metric.

Once you enable resource metering, Hyper-V will begin collecting data. You can reset metering at any time. We will then discard the data we have collected up to that point and start fresh. So, you will typically measure the utilization first, and then reset. When you measure, you are measuring the utilization over the timespan since you last reset metering. Metering is designed to collect this data over long periods of time. If you need greater granularity, you can look at performance counters. 

Having resource metering enabled and just capturing utilization data per your billing cycle has no noticeable performance impact. There will be some negligible disk and CPU activity as data is written to the configuration file.

You can try this all out for yourself now, with Windows Server 2012. In the next part, we’ll talk about how to actually use resource metering with our PowerShell cmdlets.

We hope this is useful for you. Please let us know how you’re using it! Thanks!

Question: Configuring Hyper-V Resource Metering

Windows Server 2012 Hyper-V contains a resource metering mechanism that makes it possible to track system resource usage either for a virtual machine or for a collection of virtual machines. Doing so can help you to keep track of the resources consumed by virtual machine collections. This information could be used to facilitate chargebacks (although Hyper-V does not contain a native chargeback mechanism).

Resource metering is not enabled by default. You can enable resource metering through PowerShell by entering the following command:

Get-VM <virtual machine name> |  Enable-VMResourceMetering

By default, Hyper-V collects resource metering statistics once every hour. You can change the collection frequency, but it is a good idea to avoid collecting metering data too frequently because doing so can impact performance and generate an excessive amount of metering data. If you want to change the collection frequency you can do so by using this command:

Set-VMHost –ComputerName <host server name>  -ResourceMeteringSaveInterval <HH:MM:SS>

As you look at the command above, you will notice that the collection frequency is being set at the host server level. You cannot adjust the frequency on a per VM basis. You can see what this command looks like in figure 1.

[Click on image for larger view.]Figure 1. You can change the resource metering collection frequency.

When you enable resource metering, there are a number of different resource usage statistics that are compiled. These statistics include:

  • The average CPU usage (measured in MHz)
  • The average physical memory usage (measured in MB)
  • The minimum memory usage (measured in MB)
  • The maximum memory usage (measured in MB)
  • The maximum amount of disk space allocated to a VM
  • The total inbound network traffic (measured in MB)
  • The total outbound network traffic (measured in MB)

The easiest way to view a virtual machine’s resource usage is to enter the following command:

Get-VM <virtual machine name> | Measure-VM

This command will display all of the available metering data for the virtual machine that you have specified.

Similarly, resource metering data can be displayed for all of the virtual machines that are running on the host server. If you want to see monitoring data for all of the virtual machines, you can acquire it by running this command:

Get_VM | Measure-VM

You can see what the output looks like in figure 2.

[Click on image for larger view.]Figure 2. This is what the resource metering output looks like.

Often times administrators are interested in specific aspects of resource consumption. For example, if a particular host server had limited network bandwidth available then an administrator would probably be interested in seeing the amount of network traffic that each virtual machine was sending and receiving. Conversely, if that same server had far more processing power than what would ever be needed by the virtual machines that are running on it, then the administrator probably would not need to monitor the average CPU usage.

Although you cannot turn data collection on or off for individual statistics, you can configure PowerShell to display only the statistics that you are interested in. The key to doing so is to know the object names that PowerShell assigns to each statistic. You can see the object names by entering the following command:

Get-VM | Measure-VM | Select-Object *

The column on the left side of the output lists the names that PowerShell uses for the individual statistics. You can see what this looks like in figure 3.

[Click on image for larger view.]Figure 3. You can get the object names from the column on the left.

There are a couple of things that you might have noticed in the figure above. First, there are more objects than what are displayed by default. Second, there are more objects than what I listed earlier. The reason for this is that these screen captures came from a server running Windows Server 2012 R2 Preview. Microsoft is extending the Resource Metering feature in Hyper-V 2012 R2 to include additional metering objects. In this article however, I only listed the objects that are available today.

With that in mind, let’s suppose that you only wanted to list the maximum memory consumption for each VM. You could do so by using this command:

Get-VM | Measure-VM | Select-Object VMName, MaxRAM

You can see the output in figure 4. Keep in mind that you can adapt this command to display any combination of objects that you want.

[Click on image for larger view.]Figure 4. PowerShell can display specific resource metering data.

As you can see, resource metering is useful for tracking resource consumption. It can also be useful for performing chargebacks, although there is no native Hyper-V chargeback mechanism.

Question: Resource Metering in Hyper-V

Why is Hyper-V Resource Metering Important?

The American National Institute of Standards and Technology gives us one of the best definitions of a cloud in their Special Publication 800-145, entitled “The NIST Definition of Cloud Computing.” In this document they describe a cloud as having five essential characteristics. One of the traits that they describe as being necessary to have a cloud is a measured service:

Cloud systems automatically control and optimize resource use by leveraging a metering capability at some level of abstraction appropriate to the type of service (e.g., storage, processing, bandwidth, and active user accounts). Resource usage can be monitored, controlled, and reported, providing transparency for both the provider and consumer of the utilized service.

What does this mean? A cloud enables a tenant to consume just what they need and pay for what they use. The cloud must be able to measure that usage. Using this information, a cloud vendor can charge the tenant for their resource usage.

Hyper-V resource metering

Resource metering is an important part of any Hyper-V deployment. (Image: Dreamstime)

That’s fine for a hosting company. What about in a private cloud, that is, an infrastructure that runs on premise? Traditionally the IT department is run as a cost center, or as the board of directors unfortunately see it, as a budgetary black hole. IT can change this incorrect perception in one of two ways:

  • Cross-charging: Every department that consumes IT services and resources will be given their own IT budget to spend with the IT department. IT will provide those services and resources, and invoice their internal customers on a regular basis in a non-profit manner. This changes IT into a service organization.
  • Show-back reporting: Many organizations will never consider changing IT into a chargeable service. However, IT can show the business leaders the cost of providing services to each of the business groups by reporting usage and translating that usage into a monetary value. This is a company politics move that can change the perception of IT within the business.

In my opinion, these sorts of actions could misfire and lead to talk of out-sourcing and off-shoring, so be careful!

What Information is Collected?

Resource metering will collect the following data for each enabled virtual machine:

  • Average CPU usage in MHz
  • Average physical memory usage
  • Minimum physical memory usage
  • Maximum memory usage
  • Maximum physical amount of disk
  • Total incoming network traffic
  • Total outgoing network traffic

Note that in the case of dynamic virtual hard disks, the potential size, not the actual size, is reported for maximum physical amount of disk. Also note that all data is stored with the virtual machine and moves with the virtual machine as it migrates between hosts.

Enabling and Using Hyper-V Resource Metering

Resource metering is made available using PowerShell. You can write scripts using PowerShell or you can use other tools to leverage the functionality.

You must first enable metering on a per-virtual machine basis. The following snippet will enable metering on all virtual machines on a host:

PowerShell

1Get-VM -ComputerName Demo-Host2 | Enable-VMResourceMetering

Tip: Remember to enable metering on any virtual machine created afterwards because the above cmdlet will only affect existing virtual machines.

By default, resource metering will collect metrics every hour. This is based on a per-host setting called ResourceMeteringSaveInterval. You might want to change this setting to match your billing rate in a cloud. If you are just testing resource metering, then you might want a more frequent collection. This example will change the setting to every 10 seconds:

PowerShell

1Set-VMHost –ComputerName Demo-Host2 –ResourceMeteringSaveInterval 00:00:10

After the resource metering interval has passed, you will want to collect some metrics. Here’s a quick way to see all collected data:

PowerShell

1Get-VM -ComputerName Demo-Host2 | Measure-VM
Resource Metering tracks several different metrics for Hyper-V virtual machines

Measuring resource usage by Hyper-V virtual machines. (Image: Aidan Finn)

You could do something more targeted:

PowerShell

1Measure-VM -ComputerName Demo-Host2 -Name VM01

You can get a breakdown of bandwidth usage using:

PowerShell

1(Measure-VM -ComputerName Demo-Host2 -Name VM01).NetworkMeteredTrafficReport

Armed with this information, it won’t take you too long to find resource hogs on your hosts:

PowerShell

1Get-VM -ComputerName Demo-Host2 | Measure-VM | Sort-Object -Property AverageProcessorUsage -Descending | Select-Object -First 3 –Property VMName,AverageProcessorUsage
Identify top offenders on your Hyper-V hosts using Resource Metering

Reporting on the top resource consumers. (Image: Aidan Finn)

Resource metering is a tool that can show the value of IT to the business or enable a service provider to earn revenue. And with this data, you even have some ability to track usage for diagnostics reasons.

Question: HOW TO USE RESOURCE METERING IN HYPERV

IT Professionals need tools to track usage from specific business units. If you search you can find lot of monitoring tolls that can do this job but most of them need to pay and for free open source need an advance knowledge to install, configure and enable measure metrics for the resources that you want. 

I don’t want to say don’t use Monitoring Tool in your enviroment. But it takes times and need people to do this. If you are alone you need a quick solution until decide your Monitoring Solution in your enviroment. 

So today in this article i will show you another feature that was introduced in Windows Server 2012 Hyper-V that isn’t immediately obvious and is driven by using Windows PowerShell. I will explain only basic commands that can use it every day to measure metrics of your VM’S. The feature is amazing and it’s sure that i will come back with more advance commands of Resource Metering.

Resource Metering  expose the metrics for the following resources used by a virtual machine during a specific period of time:

  • Average CPU usage, measured in megahertz over a period of time.
  • Average physical memory usage, measured in megabytes.
  • Minimum memory usage (lowest amount of physical memory).
  • Maximum memory usage (highest amount of physical memory).
  • Maximum amount of disk space allocated to a virtual machine.
  • Total incoming network traffic, measured in megabytes, for a virtual network adapter.
  • Total outgoing network traffic, measured in megabytes, for a virtual network adapter.

 Let’s start to explain with practise.

  • As usuall open a Powershell as Administrator always.
  • First we must enable Resource Metering in VM. So Type
    Enable-VMResourceMetering –VMName WIN2012X64

  • If you want to verify that Resource Metering is enable in the VM TYPE
    Get-VM –VMName WIN2012X64| Format-Table Name, ResourceMeteringEnabled

  • Let’s see the resource metrics that we get from the VM
    Measure-VM –VMName WIN2012X64

  • Let’s see more details for the VM.
    Measure-VM –VMName WIN2012X64 | Format-List

  • If you want to measure network traffic 
    (Measure-VM –VMName WIN2012X64).NetworkMeteredTrafficReport

This is my last article for the 2015. I will come back with new Articles and Tutorials in 2016.

Merry Christmas and a Happy New Year!!!!!

Question: WINDOWS SERVER 2012 HYPER-V: VM RESOURCE METERING VIA POWERSHELL

With Windows Server 2012 Hyper-V Microsoft introduced a new feature in Hyper-V called Resource Metering which allows you to measure the usage of a virtual machine. This allows you to track CPU, Memory, Disk and network usage. This is a great feature especially if you need to do charge back or maybe even for trouble shooting.

Last week I had the chance to test and implement this feature for a customer.

First you can check the available PowerShell cmdlets for Hyper-V or for the the commands which include VMResourceMetering.

123Get-Command -Module Hyper-V Get-Command *VMResourceMetering*
Get-Command VMResourceMetering

The resource metering has to be enabled per Virtual Machine. This is great, so even if you move the virtual machine from one Hyper-V host to another you still have the usage data.

To enable the resource metering you can use the following cmdlet. In my case I enable VM Resource Metering for my VM called SQL2012.

1Get-VM SQL2012 | Enable-VMResourceMetering
Enable-VMResourceMetering

With the cmdlet Measure-VM you can get the statistic for the VM.

123Measure-VM -VMName SQL2012 Get-VM SQL2012 | Measure-VM | select *
Measure-VM

To get the network traffic use the properties of the NetworkMeteredTrafficReport.

1(Measure-VM -VMName SQL2012).NetworkMeteredTrafficReport
Measure-VM NetworkMeteredTrafficReport

Here is another great thing, if you want to measure Network from or to a specific network you can use VM Network Adapter ACLs to do so. With ACLs you can not just allow or deny network traffic, you can also meter network traffic for a special subnet or IP address.

1Add-VMNetworkAdapterAcl -VMName SQL2012 -Action Meter -RemoteIPAddress 10.10.0.0/16 -Direction Outbound
add-vmnetworkadapteracl measure-vm

Of course you can reset the statistics for the VM.

1Get-VM SQL2012 | Reset-VMResourceMetering
reset-vmresourcemetering

And to disable resource metering for the VM use:

1Get-VM SQL2012 | Disable-VMResourceMetering

I think this is one of the great new features of Windows Server 2012 Hyper-V which gets not a lot of attention but is really important.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s