The "CPU Steal Time" Metric in Unix/Linux Virtual Machines and a Windows Counterpart

I haven't posted in a while; been busy both studying for Windows Server 2012 stuff and also preparing for a possible slight career shift.  But I do want to put this up here, because it's one of my answers to a Serverfault  question that I'm a little proud of.  Nevertheless, it's a deep enough topic that I expect someone who knows more about it than me to come along and correct me.  Which I welcome.  That's how science works.  I'm not learning if I'm not wrong.

Here was the question:

In order to assess performance monitoring accuracy on virtualization platforms, the CPU steal time has become an increasingly relevant metric - see EC2 monitoring: the case of stolen CPU for an instructive summary in the context of Amazon EC2 and IBM's paper on CPU time accounting for a more in-depth technical explanation (including illustrations) of the concept:

Steal time is the percentage of time a virtual CPU waits for a real CPU while the hypervisor is servicing another virtual processor.

Accordingly, it is exposed in most related Unix/Linux monitoring tools nowadays - see e.g. columns %steal or st in sar or top:

st -- Steal Time
The amount of CPU 'stolen' from this virtual machine by the hypervisor for other tasks (such as running another virtual machine).

I've been unable to figure out how to capture the same metric on Windows though, is this possible already? (Ideally for the Windows 2008 Server R2 AMIs on EC2 and via a respective Windows Performance Counters of course.)

 And here was my answer:

Let me preface by saying that I am coming from the point of view of Hyper-V as a virtualization platform because that is where I have the most experience. Even though there may be certain tenets of virtualization, as we know it, that cannot be deviated from, Microsoft and VMware and Xen all have different strategies for how they design their hypervisors.

That's the first thing that makes your question challenging. You pose your question as if it were hypervisor-agnostic, when in truth it is not. Amazon EC2, for example, uses the Xen hypervisor, and the "CPU Steal Time" metric that you see in the output of a top command issued from within a Linux VM running on that hypervisor is a result of the integration services installed on that guest OS (or virtualization-aware tools on the guest) in conjunction with data provided by that specific hypervisor.

First off let me just answer your question straight up: There is no way to see from inside a virtual machine running Windows how much time the processors belonging to the physical machine on which the hypervisor runs spends doing other things, unless the particular virtual tools/services or virtualization-aware tools for your particular hypervisor are installed in the guest VM and the particular hypervisor on which the guest is running exposes that data. Even a Windows guest running on a Hyper-V hypervisor will not have immediate access to information regarding the time spent that the physical processors on the hypervisor were doing other things. (To quote voretaq7, something that "breaks the fourth wall.") Even though Windows client and server operating systems running as virtualized guests in Hyper-V with the correct integration services/tools installed make use of "enlightenments" (which are literally kernel code alterations made especially for VMs) that significantly increase their performance in using the resources of a physical host, the bottom line is that the hypervisor does not have to give any more information to the guest OS than it wants to. That means the hypervisor does not have to tell a guest VM what else it is doing besides servicing that VM... unless it wants to. And that information about what else the physical processors are doing is necessary for deriving a metric from the perspective of the VM such as "CPU Steal Time: the percentage of time the vCPU waits for a physical CPU."

How could the guest OS know that, if it didn't even realize that it was actually virtualized? It's like The Truman Show... for computers.

In other words, without the right integration tools installed on the guest, the guest OS won't even know that its CPU is actually a *v*CPU. It won't even know that there is another force outside of itself "stealing" CPU cycles from it, therefore that metric will not exist on the guest VM.

That's why I don't even like the phrase "CPU Steal Time." The word steal just puts everybody in the wrong frame of mind from the get-go.

A hypervisor such as Hyper-V does not give guests direct access to physical resources such as physical processors or processor cores. Instead the hypervisor gives them vDevs - virtual devices - such as vCPUs.

A prime example of why: Say a virtual machine guest OS makes the call to flush the TLB (translation look-aside buffer) which is a physical component of a physical CPU. If the guest OS was allowed to clear the entire TLB on a physical processor, that would have negative performance effects for all the other VMs that were also sharing that same physical TLB. In the case of Windows, that call in the guest OS is translated into a "hypercall" or "enlightened" call which is interpreted by the hypervisor so that only the section of the TLB that is relevant to that virtual machine is flushed.


(Interestingly, that hints to me that guest VMs that do not have the proper integration tools and/or services could have the ability to impact the performance of all the other VMs on the same host, but that is completely outside the scope of this topic.)


All that to say that you can still detect in a Hyper-V host the time that a virtual processor spent waiting for a real processor to become available so that it could scheduled to run. But you can only see that data on a Windows Hyper-V hypervisor. If it is possible to see this in other hypervisors, I urge others to tell us how to see this in that hypervisor and also if it is exposed to the guests. And that is before we even get to whether that data is exposed to the guest OS or not.

My test machine was Hyper-V Server 2012, which is the free edition of Server 2012 that only runs Core and the Hyper-V role. It's effectively the same as any Windows Server 2012 running Hyper-V.

Fire up Perfmon on your parent partition, aka physical host. Load this counter:

Hyper-V Hypervisor Virtual Processor\CPU Wait Time Per Dispatch\* 

You will notice that there will be an instance of that counter for each virtual machine on that hypervisor, as well as _Total. The Microsoft definition of that Perfmon counter is:

The average time (in nanoseconds) spent waiting for a virtual processor to be dispatched onto a logical processor.

Obviously, you want that number to be as low as possible. For computers, waiting is almost never a good thing.

Other performance counters on the hypervisor that you will want to investigate are Hyper-V Hypervisor Root Virtual Processor\% Guest Run Time, % Hypervisor Run Time, and % Total Run Time. These counters provide you with the percentages that could be used to determine facts such as how much time the "real" processors spend doing things other than servicing a VM or all VMs.

So in conclusion, the metric that you are looking for in a guest virtual machine depends on the hypervisor that it is running on, whether that hypervisor chooses to provide the data about how it spends its time other than servicing that VM, and if the guest OS has the right virtualization integration tools/services/drivers to be aware enough to realize that the hypervisor is making that data available.

I know of no way on a Windows guest, integration tools installed or not, to see how much time, in terms of seconds or percentage, that VM's host has spent servicing it or not servicing it respective to the total physical processor time.

EventLogClearer v1.1.3.22

I have released an updated version of my EventLogClearer, bringing it up to version 1.1.3.22. For the original release, see this post.

EventLogClearer 1.1.3.22

Improvements made in this version include:

  • Fixed a bug where the application acted weird if you ran the log clearing procedure two or more times in a row.
  • Added a new mechanism for supplying alternate credentials, instead of only being able to run as the currently logged on user. This applies to both auto-populating the list of computers from AD, and running the event log clearing procedure. If you leave the credentials blank or as the default, "username," the current user will be used.
  • Added the ability to clear a ton more Applications and Services logs than before, due to me realizing the potential of the EventLogSession class.

As before, .NET 4.5 is required to run the application. The project was built in Visual Studio 2012.

Here is the executable: EventLogClearer-1.1.3.22-exe.zip (68.71 kb)

Here is the source code: EventLogClearer-1.1.3.22-source.zip (308.11 kb)

Sometimes I Can Access the WebDAV Share, Sometimes I Can't!

You probably already know that all of the Sysinternals tools, such as Process Monitor, Process Explorer, Autoruns, and much more, can be accessed via "shared folder" from any computer connected to the internet by navigating to \\live.sysinternals.com\.  This isn't the same kind of share you'd create if you just shared a folder on your PC.  It's a WebDAV share, and is accessed over HTTP.

Sometimes though, I feel the need to access this share from the command line, either in the Cmd shell or Powershell.  Alas, here's what I see:

Network path not found*Path not found.*

I get the same result with Powershell. Bummer. Well I know I can access the path with Explorer when I type that same UNC into the address bar, or if I just type the UNC into the Run dialog box, so this must just be a limitation of those command-line tools, right?

It works in Explorer*Works fine in Explorer*

Oh well... but wait. Now having successfully accessed the network path with Explorer, let me now immediately go back to the Cmd shell and try it again:

 

Now it works in Cmd too!*Now it works in Cmd too!*

OK, now accessing the network path works fine from the Cmd shell and from Powershell, even though all I did was access it through Explorer first, and then try again. Now I just have to know what the heck is going on... and to do that, I need to use Process Monitor. Which, amusingly, is in the WebDAV share I'm trying to access. But I'll run a local copy.

I started the trace. Here's my first attempt to access the network path with Cmd.exe, which failed:

Cmd.exe network path not found*Network path not found*

This was the very first time in the Process Monitor trace when the string "live.sysinternals.com" appeared in the Path field. It's also the first time the Cmd.exe process shows up in the trace. It's currently filtered to only include events where the Path field contains the string live.sysinternals.com. The really interesting part about this is that it appears the moment I pressed Enter on the command line, Explorer.exe was the first process to be involved, not the process I was interacting with! That's odd. Maybe a file system filter driver intercepted the call and notified Explorer? It looks like Explorer is looking for something related to named pipes and the Workstation Service (wkssvc) on the remote server, but it doesn't find it.  Then Cmd.exe first checked my local file system for a file in Windows\CSC\ directory, which it didn't find, and then it tried to access the network path that I actually asked for, which resulted in "Bad network path." Then it apparently tries again with the same local file system path, and then again with the network directory instead of the specific executable name.  All failed. "Network path not found," my command prompt tells me. But with no further input from me, Explorer takes off doing its own thing, calling cscapi.dll and loading things in the background and sending things over network. All I did was hit enter in the Command Prompt above.

So what is this CSC directory? Googling the term led me to an old post on Raymond Chen's blog. Client Side Caching. OK, so apparently both processes are looking for a cached or offline version of the network path.

Then I move over to the Explorer.exe window and type the path into the address bar. Explorer looks for some more CSC stuff first, and then svchost.exe starts communicating with the remote server over TCP. There's a lot of loading of WebDAVRedirector stuff. Finally, after a lot of work, I start seeing events like these from Explorer:

Explorer finds it, finally*Explorer starts finding it, finally*

Notice that Explorer also seems to be storing the autoruns executable in a temporary "Tfs_DAV" directory on my workstation.

Finally, after having success with Explorer, I go right back to the Command Prompt and try it again. This time, the trace looks like this:

Works in cmd.exe now too

Now I see svchost.exe stepping in with a WebDavRedirector, and cmd.exe getting some successful returns from its IRPs. Finally, after playing around in that Tfs_DAV directory and some more intermingling of svchost.exe and the System process both helping out, the process autoruns.exe finally launches.

So that's a pretty fast and loose overview of what is actually going on. The entire trace was a beast to wade through, and there is obviously a lot of orchestration and cooperation required between many different Windows components required to allow you to access a WebDAV share from within Cmd.exe and I don't fully understand all of it... but the bottom line is that at least on my Windows 7 SP1 x64 workstation, it looks like Explorer.exe is smart enough to read from a WebDAV share and cache the data locally, whereas Cmd.exe is only smart enough to read the data locally, if and only if it's already cached locally... or perhaps the redirector had to be "woken up" by Explorer first, before Cmd.exe was able to use it.

Finally, I'll leave off with a bit about the WebDAV Mini-Redirector from Wikipedia:

"In Windows XP, Microsoft added the Web Client service is also known as the WebDAV mini-redirector[11] which is preferred by default over the old Web folders client. This newer client works as a system service at the network-redirector level (immediately above the file-system), allowing WebDAV shares to be assigned to a drive letter and used by any software. The redirector also allows WebDAV shares to be addressed via UNC paths (e.g. http://host/path/ is converted to\\host\path\) for compatibility with Windows filesystem APIs."

Log Parser 2.2 and Log Parser Studio

At first I thought to title this post the same as the catchphrase of Log Parser: "The Whole World Is Your Database!"

But then I decided that was a bit too exciting for what I actually wanted to talk about.

So I just discovered Log Parser Studio a few days ago. LPS is a graphical frontend to Log Parser; quite similar to how SQL Management Studio is a GUI frontend to interacting with SQL Server.  I am, quite frankly, ashamed that I didn't already know about it. It's fantastic.

The thing is... Log Parser is a command-line utility that uses a very SQL-esque language to interact with logs. What kind of logs, you ask?  Any kind of logs! That's right... you can use it to query the Windows Security Event Log, or you can use it to query a folder full of IIS web server logs, or you can use it to query a log full of your own personal electric utility bills from last year!

However, Log Parser itself is a very complex, albeit powerful and flexible, command-line utility. Maybe you want something a little more user-friendly to get you started. That's exactly where Log Parser Studio, the GUI frontend, comes in to play.

As a little demonstration, I installed Log Parser 2.2 on my workstation. Then I downloaded Log Parser Studio to my workstation. I fired it up as a Windows application, and I pointed it to the remote IIS logs directory of this very web server. I then right-clicked on "IIS: Request per Hour" and chose "Run report now." As if I had just run a SQL query in SQL Management Studio, this window popped up:

 

Log Parser Studio Query*Click for Bigger*

This data is probably every single HTTP GET request made per hour, rather than a count of hits made by unique IP addresses, but the point is you now have this amazing utility that will parse practically any amount of data you can think of from any source of data you can think of. Go check it out and see how Log Parser is even capable of generating pie charts and bar charts and all sorts of crazy things using this data!