Microsoft Azure AZ-801 — Section 8: Implement and manage Storage Spaces Direct Part 4

52. Modify quorum options

Let me show you how we can configure our quorum settings.

Now, the quorum is the capability that’s going to manage votes and all of that. And it’s actually pretty easy to configure even though they’ve kind of hidden it from you.

So, here I am in failover clustering. If you forgot how to get there, all you got to do to get there is just go into Server Manager and go Tools and you’ll see failover cluster manager. So, from there where they’ve kind of hidden this is if you right click the cluster namespace here, which is NYC-CLUSTER.examlabpractice.com. In my case, if I right click that there’s an option here it says More actions and that is where you can configure your cluster quorum settings. So, we’ll click on that and then we’re going to click Next and you’ll see what the default is here. They’ve literally the default option is use default quorum config, which is basically to do dynamic quorum, which essentially means that it’s going to make a decision whether it should use a witness or not. If you’ve got an even number of servers, it’s going it’s going to going to go with a disk witness. And if you’ve got file share available, try to do a file, share witnesses, various things there that it’ll try to do. But in my case, because I’ve set up an iSCSI setup, it’s going to use my ESXi as a disk witness. But you can say select the quorum witness if you want. You can click Next. And then if you want to manually configure a disk witness, you can if you want to configure a file share witness, you can do that. All you got to do to configure a file share witnesses go through this process of creating a file share.

Now, let me warn you file share witnesses. The file share would have to be hosted on a different server. It cannot be hosted on the same server that is part of the cluster. So, you’d have to have another server to do that with.

 So, I can’t really demonstrate that just because I’ve only got two servers here, but that would be the deal. And you just basically go through this process of, of creating a share. You can, like I could say, NYC-SVR1. I can’t really do it on Server1 because Server1 is part of the cluster. But just to show you an idea, I could I could set the share name to like witness, and then what you would do is you go to the custom permissions and you would give you go up here to options or objects, computers, and you would give NYC. Server1 full control. You would give NYC-DC1 full control and you would also want to make sure you gave yourself domain admins, group or yourself explicitly full control. In order for that to work. You could also remove the everyone group and that’s how you would set that up. Again, I can’t do that because I can only do that if I have another server that’s not part of the cluster. But that’s the process. I can also do a cloud witness, which means I could tie it to Azure, but you have to have storage account and all that. And not getting into that in this video. Or, I could say do not configure the quorum.

So, those are my options there now can go back and there’s also an advanced quorum config and all that really let you choose what’s going to get a vote. So, if you had a bunch of servers and you wanted, you know, maybe, one or two of the servers not to be part of the quorum, you can do that if you want. So, that’s really all that is.

So, that is how we can configure our quorum settings in our failover cluster.

53. Configure network adapters for failover clustering

Let’s talk about your network adapter configuration for failover clustering now. In a perfect world, if you’re dealing with physical servers, you definitely would want to have multiple adapters in your server. And the way you want to think about this is that, you know, you’ve got information that’s being sent back and forth between your different servers that are part of the cluster that are like heartbeat type information that let the servers know that they’re still up and running. And you want to you want to decrease the latency of that as much as you can so that there’s not any kind of delay or any of that. Because if there is, then of course, at that point one of the servers could think, oh, well, that server is down and I need to become active or all the load needs to shift over to me. And then you run into issues there, right? So you’ve got to think about that in a very, very small environment where you’ve got good bandwidth, maybe, fibre and everything else. Then I would say, yeah, you probably can get away with just having the one network adapter, but you are running the risk of not only affecting the cluster, but the other thing is what if that network adapter fails in that machine, right? And that’s where, of course, something called NIC Teaming comes into play. Like, like here I am on NYC-DC1 and. You know, I can go over here to the left and go to local server and you can see currently I don’t have NIC Teaming on my machine.

Now, if I had multiple nicks, I could. I can go here to disable where click on the disabled link and I could add my multiple NICs here and I could link them together and that would provide load balancing and speed performance for my NICs. In fact, in bigger environments, companies will have for NICs in their server, they’ll have two. NICs dedicated to regular network traffic and they’ll have two. NICs dedicated to the to the cluster information. All right. And so from there, what you would do is if you had multiple NICs in your machine, you can go Tools, you can go to Failover Cluster manager, and then from there you’ll click down here on this networks area here. So, you click on that and then if you go to properties, you can tell it. So, it would detect that there’s multiple NICs here, you can tell it which adapters here. I’m going to support this on right now. Is allow cluster network communication this network because it’s detected that network.

So, in a perfect world, you would want to configure a separate network for your cluster, and that can be nothing more than just a separate switch that in the real world that your equipment is plugged into and you would have that cluster information going back and forth for that, for the heartbeat information and all of that’s going to happen on the cluster. In that case, if you’ve got the multiple network networks detected, then you could switch over to that other network. Right. And you could say do not for this particular network. If this is the network that clients are on, then you could say do not allow cluster communication that. And that’s how you can configure those network adapters for failover clustering.

So, here I am in Hyper-V manager on my host computer and I’ve got my two servers obviously that have failover clustering. And one thing I could do is I can go here to Virtual Switch Manager and I could just create a private network switch.

Now, to do that, I just go up here to New I choose private, I click create, give it a name and a way I go, all right, I’ve already got one created. And then what I would then do is let’s go over here to this DC1. I’ll right click and go to Settings and then we’ll go here and choose network adapter or sorry, add hardware, network adapter, and I can just add another adapter and connect it to the private network.

So, we’ll go ahead and do that and click Okay on that. And then I’m going to do the same thing on this server here.

So, we’ll go add hardware, network adapter, Add. And then private network, we’ll click Okay. From there, we can jump back into our servers here. And let’s just close out of this tool here. Here we go. And close out of all of this. All right. These right click network settings here down in the lower right. I’m just waiting on that to load up. All right. And as you can see, it’s trying to identify this new network that it’s detected here. And I can click change it after. Here it is. I’m going to right click. I’m going to go to properties. TCP IP4 and let’s just give it an address. Let’s say 10.0.0.1. All right. And DNS will just be 10.0.0.2. Don’t really need DNS for this anyway, but we’ll click Close.

Now, we’re going to jump over to DC1. We’re going to do the same thing. Just close out of failover clustering right now. Close out of all of this right Click Adapters, Open network and Internet settings. Wait on that to load up. We’ll go to Ethernet. I’m going to change adapters. Right click our second one here. TCP IP4. We’ll give this 10.0.0.2. 255.255.255.0, and DNS will be 10.0.0.2. So, we’ll click Close. All right.

Now, what we can do is to go to Start, to Server Manager. We got to wait on the little blue bar to quit spinning here. This is just letting the services have time to properly be detected in Server Manager. So, that’s what it’s doing. It’s just querying these different services. Okay, probably enough time. Let’s go Tools and go to Failover Cluster Manager. Okay, so failover cluster manager is loading up here. Slowly but surely, we’ll expand that out. We’re going to go to networks. And notice that a second network has been detected right here.

So, now what we can do is we can go to properties on that. And we can say do not allow cluster network communication this network. And so says the network is no longer available to clients to connect cluster roll. That’s fine. Click Okay. So, now this is going to contain our cluster.

So, that’s how you would do that in a virtual environment. Obviously, in a physical environment where you have physical servers, you’d have to set up the equipment and all that and connect together. But hopefully now that gives you a better visual on configuring the network adapter settings for your cluster.

Now, real quick, because of some of the demonstrations that I’m going to do, I’m actually going to allow this network right here to be to accept cluster information. So, I am going to go back to that and we are going to allow Network 1 to accept communication as well. So, just I don’t know if you’re doing this with me or not, but if you are, that is what I’m doing. All right. And that’s because some of the demonstrations I’m going to do is going to involve Network 1 needing to be connected.

54. Configure cluster workload options

Let’s talk about providing redundancy now for various workloads involving clustering. And so what I’m going to start off here doing in NYC-DC1, I’m going to open up Server Manager and I’m going to go to manage add roles and features and I am going to install. Let’s go right here. And let’s choose DHCP service. All right. Go ahead and install that. I’m going to jump over to Server1. I could actually have done this remotely, but I’m just going to do it separately.

So, from there, we’ll go start. Server Manager. And we’re going to do the same thing. We’re going to install the DHCP service. Okay. So, we’re just letting the little blue bar kind of get done spinning as usual. And then once that’s done, we are officially able to go to manage add roles and features. I think I probably waited long enough. We’ll see. If not, it’ll let me know.

So, here we go. Next. Server1. All right. And we’re now going to select the DHCP service. Next, next, next. And install. All right. So, we’re now installing that service over there and still installing on my NYC-DC1. So, I’ll pause the recording while that’s happening. All right. I’m just going to hit Close now and hearing him on my fellow cluster manager. And. We can go right here to Rolls. And right click that and you can click to configure roles.

So, these are the various types of workloads that we have available, right. Built right out of the gates. With clustering, we could do distributed file services, namespace server, we could do DHCP, which is kind of what I’m getting into right now. DTC, File Services. You can do generic applications if you have an application that you need to provide high availability for, there’s some configuration you can do there. Granted, there are we could probably spend 60 hours talking about the little nitty gritty of every single possible solution and scenario. When it comes to clustering with applications, you kind of have to research things like if you’re going to do application clustering, there’s lots of reading material that shows that you can even have scripts, generic scripts and provide high availability for those. You can do generic services, just Windows services if you want broad high availability for those Hyper-V iSCSI services message queuing other servers they tell you, which is kind of a funny one, but this would get in another wizard basically that would point to certain application type services that are running on the machine. You can provide high availability for virtual machines, even the good old Windows Server from back in the day, if you know what that is for net bias services. So, there’s lots and lots of options here that you can go with.

Now, the other thing I want to point out here, you know, with like DHCP and all that is that, you know, you can run through the wizard but a lot of services to have their own, their own way of managing clustering services from within the tool. And if there is a service that you can manage through the actual tool itself, it’s usually better than to use the failover wizard.

So, for example, if I come over here to, I’m going to go Tools and I’m going to open up DHCP. We’ll go right here and expand this out. And let’s just let’s just create a scope on DHP. I’m not really explaining DHCP thoroughly here. I just want to show you the failover. But if I go here, I’m just going to call this ScopeA. And we’ll create a scope. Let’s do 10.1.0.1 through 10.1.0.100. And we’ll do 255.255.255.0. So, this handout, those addresses don’t have any exclusions, not least duration, not going to set any of these options.

So, now let’s say that’s my scope, right? And what I can do is I can right click that scope and you’ll notice there’s a configure failover option here. So, I can go there, select that, and I can choose, let’s say in CDC, one is our primary server, so we’ll just say NYC-DC1. We’ll click and let’s see, following scopes already exist on the server. Oh, I’m sorry. I chose. I meant that was the partner server. Sorry, not primary server remembering go. All right, there we go. NYC-SVR1. There we go.

So, we’ll add that server. We’ll click Next. And we’re setting up our relationship to do a 5050 split on the IP addresses, which I’m not getting into all this right now because this is just getting into the intricacies of IP splitting right now. I just want to show you the failover being established here. All right.

So, I can choose if I want to do a password between the two, I can. I’m not going to I’m going to click Next and then finish. And it looks like it was successfully set up. Let’s jump over to NYC-SVR1 and see how that’s shaping up. We’ll go Tools. And we’ll go to DHCP. And let’s see if the scope is appearing. And it is so our scope is appearing, of course. Indeed, as far as DHB goes, you would have to activate the scope before it’s going to hand out any addresses. But right now, if these were both activated and NYC-DC-1 was to fail or whatever, it would fail over to Server1.

So, the thing that get across here, though, is that you’d have to kind of figure out what kind of workload you’re talking about, what kind of role you’re looking at here. And does that role have its own service tool for managing the failover? If it does great, you don’t have to use the failover wizard for it. Exchange is a good example of that. SQL is a good example that there’s a lot of services out there that have you’ll have to put the fill over service. You have to install the fill over services and stuff on your server, but then you can use those tools to actually get the, the workload, the role service working. All right. And so the big thing there, though, that’s the main thing I want to get across to you. You need to figure out, “Well, in my real world environment, what service am I trying to provide high availability for?” Do a little research on it. You know, you’re going to use that fill over service to do a little research on how we can configure it.