Time
14 hours 13 minutes
Difficulty
Intermediate
CEU/CPE
20

Video Description

Virtual Machine Performance Monitoring Lab This lesson discusses different methods of monitoring virtual machine performance. In this lab based class, participants learn about using production resource pools as a way of monitoring virtual machine performance. Participants also learn step by step instructions in setting up a chart to see a virtual machine's performance.

Video Transcription

00:04
Okay. Welcome to lab number 17. Where would we be looking at the
00:09
different ways to March or the performance of our virtual machines
00:13
from the slab. We still have our CPU busier running.
00:17
I've put the resource pulls back to their original configuration.
00:22
So the production resource pool
00:25
is that the high value of 8000
00:27
chairs?
00:29
And we can see that
00:32
the amount of host member used by the production,
00:35
uh,
00:36
resource pool and the guest member used
00:40
1.35 Gig 1.25 gig. Here
00:43
I go to my test
00:45
resource pool. I've got low shares 2000
00:49
and I've got much more host memory being used and more guest memory
00:54
to this is because of the
00:56
The additional constraints on the CPU for
01:00
resource is for that
01:00
VM. So the memory usage looks a little different between the two resource pools.
01:08
Okay, so
01:11
what we're going to do now is go into the Web client
01:15
and
01:17
I'm actually gonna open up
01:19
to Web clients.
01:21
Let's put one there,
01:23
and then we will
01:27
open. Another
01:30
should automatically log me in.
01:38
It's taking a moment.
01:41
Okay,
01:42
So have his window side by side. So we can see
01:46
the different performance characteristics of
01:49
of each of these V m.
01:53
Okay, so let's go to Vienna's and templates.
01:57
I've got a
02:02
make this a little bit larger
02:08
and then we'll stretch this one. Just a touch.
02:13
Okay,
02:15
so underneath the mantra tab
02:19
selecting performance,
02:21
I can then see what this particular VM
02:23
looks like.
02:25
Just like with the Vic Aiken. Select real time
02:29
updates,
02:32
and we can see that our processor is maxed out.
02:38
Do the same thing over here
02:39
except here. I'm going to look at my second VM. That way I can compare the two
02:51
go to the mantra tab,
02:53
select performance,
02:59
and I'll set this for real time as well.
03:05
And both both these v EMS have pretty similar characteristics at this point. The CPU usage is still maxed out.
03:13
All right,
03:16
now we'll go look at the advanced
03:19
performance tab or button
03:22
and this view we can see
03:24
a, uh,
03:27
another chart for CPU.
03:30
Let's make this full screen for just a little bit.
03:36
The default is CPU, so we'll leave that set there. We can look at other things, but that's where most concerned about right now.
03:43
So what I can do is go to chart options
03:46
and I've already configured this one. But normally win the name of the VM would be checked here. So I'm gonna uncheck that
03:53
and
03:54
I'm going to de select all the counters.
03:58
So these are all the things you can display in a chart,
04:01
and we're only going to select two that were interested in
04:05
will select Ready.
04:09
So this
04:10
shows the
04:13
percentage of time the virtual machine was ready but could not get scheduled to run on the physical CPU.
04:17
So that's a no indication of
04:20
that. The fact that this virtual machines waiting for its chance
04:25
two actually
04:27
execute some instructions on the host CPU
04:30
and then also select used.
04:33
And this is a total CPU usage.
04:36
Okay, so the Times fan will say real time notice. Of course, we can pick other options,
04:43
but we'll stick with real time for now,
04:45
and that will start to load that graph.
04:48
We can see we've got our
04:50
our red line and our blue line,
04:56
the ready values or not
04:58
changing very much.
05:00
But see, if you used
05:01
is
05:02
kind of hovering around that 21,000 level.
05:08
Okay,
05:11
let me just drag this tab here. That might be a little easier to see.
05:14
All right now I will configure the same chart settings for the Window seven clone
05:20
again. We'll go to the advanced button
05:29
and we'll click chart options.
05:30
You might have also noticed that I'm getting some more
05:34
alerts for my CPU usage being very high. So the center is trying to tell me that there are performance problems.
05:42
Of course, we know that we created those ourselves, so
05:46
that's to be expected. Okay, So I will do Select
05:49
machine,
05:50
I will say
05:51
De select all the counters
05:55
and then go ahead and select ready
05:58
and used.
06:09
Okay, So now what we can see is
06:13
now this is running for Ah, a few minutes. We can see that our average CP ready is just just over
06:20
six milliseconds
06:23
and
06:24
the average
06:25
ah used value
06:28
is 15,007. 22.
06:32
Now, this is for the window seven V m, which has
06:36
which is in the production resource pool which has high
06:41
CPU shares allocated.
06:43
So
06:45
six mil seconds in 15 point
06:47
7 15,050 roughly four
06:50
ready and used. If we look at
06:57
the other machine,
07:01
we can see here that
07:05
No, this chart is not configured. I thought I had this one setup.
07:09
Hold on one second.
07:12
Go ahead and
07:14
fix that.
07:16
Used
07:18
P and ready.
07:25
Okay.
07:27
So in comparison,
07:29
we can see that the,
07:31
uh
07:32
we're about a little over a millisecond. Higher average for the ready value, because the other one was just over six. Now we're at seven milliseconds,
07:42
and we've got
07:43
18,000 milliseconds vs,
07:46
um,
07:47
15,000.
07:53
So again, proving
07:55
that
07:56
because the machines are in different resource pools and have different CPU resource allocations,
08:01
we're able to directly
08:03
see the
08:05
the difference in the in the performance
08:07
and the late and see that that is being felt by
08:11
by those applications. Users would be running on those systems.
08:18
So again, that's about a millisecond faster for the production VM and the 3000 milliseconds. Faster
08:26
for the use value.
08:30
All right, so now what I'd like to do is
08:33
stop these scripts from running,
08:35
so I'm gonna go ahead and click. Stop,
08:39
stop.
08:43
And if we go back to our web client,
08:46
the next update should show that
08:48
these values are going to be much different.
08:56
I can already see the CPU usage dropping off
08:58
is what we expect.
09:03
And as another way of looking at the overall effect this has had, we can go back to our vic
09:09
and select our host.
09:11
Since I can do Performance monitor and the host,
09:15
you can see that the CPU was
09:16
was quite busy all over the place
09:20
during the time of this testing.
09:24
So we got very close to, ah, 100%
09:26
for all of the various cores that we're trying to
09:31
satisfy the requirements of the CPU busier tool.
09:35
And now that that's no longer running
09:37
RCP, you said should should go back down to something much lower because the
09:43
system is more or less idle.
09:50
Okay, so
09:52
that concludes lab number 17
09:54
and live 18.
09:56
We will be working with alarms.
09:58
And of course, you've already seen some of those alarms get triggered
10:01
or alerts are getting triggered
10:05
to show me that I've got
10:07
huh?
10:07
You know, my health status monitoring is, ah
10:11
telling me something's
10:13
and we'll have a little bit closer. Look at that. So we'll we'll create some alarms. That monitor for conditions will create an alarm that monitors for an event. And we'll talk about the difference between those things.
10:26
And then we can see what the what the alarms look like when they get triggered and how to acknowledge them and finally disable them.
10:33
All right. Thank you. See you in the next lap.

Up Next

Virtualization Management

Our self-paced online Virtualization Management training class focuses on installing, configuring and managing virtualization software. You?ll learn how to work your way around the cloud and how to build the infrastructure for it.

Instructed By

Instructor Profile Image
Dean Pompilio
CEO of SteppingStone Solutions
Instructor