Microsoft Azure AZ-801 — Section 17: Monitor Windows Server by using Windows Server tools and Azure services

100. Monitor Windows Server by using Performance Monitor

I’d like to look now at how we can use Microsoft’s performance monitor for monitoring a window server.

Now, performance Monitor has been around for 20 some odd years now. Ever since the 1990s, it’s been around and it’s kind of been a go tool when it comes to just wanting to monitor the various counters for performance. A lot of people call it performance and it’s been available pretty much in all the different Windows operating systems over the years. Let’s take a look at it now. Here we are on our server, one of our servers. We’re going to open up Server Manager and we’re going to go Tools and then we’re going to click on performance monitor.

So, once we have opened up performance monitor, we’ll go ahead and just click on performance monitor. And right out of the gate, you can see that the performance monitor is just monitoring the processor.

So, the percent processor time and it just does a count of every second. It’s showing you what the processors do. And of course, if I open a bunch of things and close a bunch of things, you’ll see it spiking up, you know, just kind of clicking a bunch of things. You can see my processor spiking up, but as I stop doing that, you’ll see it goes down. So, it is monitoring things in real time.

The real fun with all this is getting into monitoring the various counters, right? The counters are the various objects that you can actually get into and there are for each object, you’ll find there are different counters that go along with that object.

For example, if I click this little plus sign, you can see this list of objects. So, these are the different objects I can monitor. Now, under each object, there are different counters that I can look at. So, I just expanded processor and you can see these various these various counters underneath processor. And of course, you know, like most people, you’re going to look at this and you say, I don’t know what all this stuff is. What you can do is go down here and click Show description. And Microsoft does provide a description of what these counters actually do.

If I look at the amount of interrupt requests I’ve got. The amount of idle time, maybe, my processors go in through and then, maybe, I can go and look at some of the other counters as well. Like, let’s look at there’s our network adapter bytes received per second. Total bytes received total bytes sent and received per second. I could do that. And then how about there’s memory, maybe, I want to see my available megabytes of memory. You could go down and look at your physical disk. All right. The percent read time, physical disk, because there’s various counters I’ve added and then I can click, okay. And then I’ve got these different counters listed here. Oh, another thing you can do is you can actually connect into a remote computer and monitor remote computers as well. You can do that right here by putting in that computer name of another computer.

The other thing you can do is when you’re looking at these various counters, you can click this little highlight icon and it will highlight what counter you’re currently looking at. So, you can see I just clicked on the processor counter. I can see that it’s highlighting that out. All right. And you can see what the last the last bit of information the counter was with the average is what the minimum and the maximum is. And this is definitely give you an idea of if something’s too high or low, figure out what the average is right now. For example, Microsoft generally will say that anything over 80% of your processors idling over 80% all the time you’re working your process are too hard. It’s okay for processors and memory to spike up there, but it’s when you’re when you’re above 80% and it’s idling there, you’re obviously pushing this too hard. You might want to consider offloading some of the services of a server if you’re doing that. All right. Another thing we can do is come over here. We’re viewing the current activity. You can view the log data. You can actually download log data information, but you can also go here and you’re looking at a line diagram. You can switch this to histogram view if you want. And then you can also do a report which just gives you statistical information.

So, there’s various options there as far as what you’re looking at. All in all, there are hundreds of counters, obviously, and there’s no way to memorize all of them. The main thing to understand is that you can look at the description of these things and generally pay attention to the 80% rule on things like processor and all that. Anything above 80%, if it’s idling at that, idling that high, it’s obviously being pushed a little bit too hard. And you want to consider taking some of the load off of it. If you’re really if you’re really pushing it to that, that level of extreme.

Now, there is exceptions to that 80% rule. For example, like I just looked at idle time. That’s my hard drive being idle. You’ll see it’s idling above 80%. That basically means that it’s not really being utilized. It’s idle. Right.

So, you have to think about what that counter’s job is and then and then base your 80% rule on that. Okay. All right. But all in all, again, very easy to use. Performance monitor is definitely a helpful tool and something that you should play around with to get a feel for if you’re not already familiar with it.

101. Create and configure Data Collector Sets

Something else that can be really helpful when it comes to performance monitor is a feature called Data Collector Sets.

Data collector sets allow us to collect information over a period of time so I could collect statistical information over an extended period of time. This can be helpful for a couple of reasons.

Number one, it can help you establish what’s called a baseline. A baseline is where you kind of establish what’s normal for your server on a day to day basis. And then what’s great about that, if you if you save that information to understand what’s normal, establishing a baseline, then later down the road, if things are not acting like they normally do, maybe, things are slowing down on your server and things are abnormal. You can grab the same counter information as you did with your baseline and you can compare the two and you can determine, well, what’s high now versus what was high when I did my baseline.

So, if you establish a baseline by gathering this information and then later down the road, grab the same information, you can compare the two and figure out what’s abnormal. Another reason that having a data collector set is helpful is when you’re trying to prove to somebody like your boss that, maybe, it’s time to upgrade our equipment. Were slowed down. We don’t have enough space, we don’t have enough processing power, don’t have enough memory network, bandwidth, whatever it may be. By having these data collector sets, I can generate a report and I can take that report to my boss.

So, let’s take a look at that. Here we are on Server Manager or our server, we’re opening up Server Manager. You’re going to go Tools, going to go down here to performance monitor. And once we’re in performance monitor here, we have data collector set so we can expand that out. And there’s a couple of different options when it comes to that. Number one, I could create a user defined data collector set, which kind of lets me just pick and choose what counters and things I want to collect data on and also lets me kind of pick how long I want to collect it., maybe, it’s going to be a week or a month or whatever, and then there’s a couple of handy ones that are just built in like an Active Directory diagnostic one system diagnostic test and then performance.

Now, since we’re learning about performance, why don’t we pick out performance? So when I go right here to performance and click that, you’ll notice there is a little play button up here which will which I can hit, which will start the data collector set. All right.

Now, notice that when I do that, this tiny little play button appears over system performance. Now, that essentially means that this data collector set is currently collecting information. So, it’s got it’s got a bunch of different calendars that it’s now monitoring. And by default, well, with that particular data collector set that built in when it’s just going to count for 60 seconds, again, if you want to count longer, you can create what’s called a user defined one, which is a custom one. You can choose however long you want to do it.

What will happen when that’s done? I can go down here to reports and then go under system and then system perform and see how they kind of mirror each other. Under system performance, you can see this report being generated. And as you can see, it’s still collecting because it hasn’t quite been 60 seconds yet. All right. Now, once that 60-second period is complete, you’re going to see that. It’ll tell you that it is now generating the report. So, this particular report is just a 60-second period. All right.

So, there it is. You can see that it says that it’s generating the report. All right. And then as soon as that’s done generating, it’s just going to pop up on the screen and I’m going to have access to all these different statistics. All right.

So, take a look. Look at all the different statistics that I’ve got available to me involving my CPU usage. It says that the CPU is normal, green is good. That means things were pretty normal. It wasn’t being pushed very hard. But I can go under here and expand CPU out. I can expand the processes. You can see all the processes that were running during the time of the report being generated. There’s the processor, the minimum maximum mean time. I can take a look at the services that we’re running, various services that we’re running at the time. All right. The network and what the network was doing. So, IP address information, if any packets were being sent out, failures, any IPV six, v four. Lots and lots of stuff here. You can do lots of stuff you can look at.

Now, the other thing, of course, that’s helpful is that I could print this report out. So, if I, if I had to take this to my boss or something and be like, Hey, look, let me show you what’s going on with our server and why things are really slow. I could do that, right? You could also save this by printing it to like a PDF file or something. You save a digital copy of it, which is nice as well.

So, there is other options as far as you don’t have to go to using a paper based system if you don’t want to. You could, you could print it to a PDF file. So, that’s also helpful.

So, anyway, hopefully that gives you now an understanding of what data collector sets are on performance monitor.