Managing Storage in vSphere 7

6. Demo: Connect anSix Host to an iSCSI Storage Array

Here you can see I'm logged into the home screen of the VSphere client, and I'm going to go to Hosts and Clusters, and here's a new host that I've just installed in my lab environment. And on this new host, on the Configure tab under Configure, you can see the storage adapters area. And as a side note, in my lab environment I do have anDiscuss storage solution ready to go. So what I want to do on this host is add a new storage adapter. And so what I'm essentially adding here is basically a way for this Six host to connect to a particular storage device. And as you can see here, there are already some storage adapters that exist inside of this Six host. So let's take a look at the first storage adapter. This is a host bus adapter, or HBA, Vohraone. And if I click on devices, we can see exactly what loons are available through this storage adapter. And so yeah, I can click on my different storage adapters. This storage adapter was plugged into my CD-ROM.

This storage adapter is connected to the local data store on this Six host. So that's one type of HBA that connects to these local discs on this Six host. But now I'm going to add a new storage adapter, and this is going to be a software adapter. And you can see here I have a few options. I can add a software Izzy adapter, I can add a software fiber channel over Ethernet adapter, or I can discover NV Me or RDMA adapters. I'm going to go with my first option here. I'm going to add a software Isuzu adapter. So I hit OK and we should see it show up momentarily here. There it is Vohra 65. That's my new softwareIsguzzy adapter, and we can see that an Incas been dynamically assigned to that adapter. The IQN is a unique identifier for Izzy devices. As a result, every Ischia device, whether it is a software initiator, a hardware initiator, or an actual storage system, has these unique IQNs that allow them to be identified in the Issyk environment.

So yeah, now I can click on my new adapter. I can see the properties of the adapter down here. So what I want to do next is configure the software’s scuzzy initiator to communicate with my eyes. Scuzzy storage apparatus I wanted to learn about lenses and the available lenses that exist on this scuzzy storage device. I also wanted to learn about existing data stores as well. So I'm going to go to Dynamic Discovery and I'm going to click on Add. And so here I'll go ahead and put the address of my scuzzy storage array. In this case, it's 192-168-1994, and I'll go ahead and click on OK here and look at what it's telling me. Due to recent configuration changes, anare scan of this HBA is recommended. What a rescan does is issue a Send targets request. So when I rescan this adapter, what's going to happen is my Six host is going to reach out to this storage device and discover the available lunge on that storage device. So right now, that adapter is being rescanned. I can see the progress here under recent tasks, and it looks like it's completed. So let me click on devices here and refresh my list of devices here. And I'm also going to rescan storage as well. Search for new VMFS volumes on this Six host. And at this point, we can see that under devices, nothing is showing up. So what we should be seeing here is a list of all of the ones available on this ice sculpture storage system, but we're not getting anything back.

And so something has gone wrong with this process here. So what has happened at this point is yes Xi host has sent a request to this storage device saying "Tell me about your lungs" and it's not getting anything back in return. Now I know that I have a working icecussystorage device, but maybe that sczzy storage device is not configured to communicate with this particular host. Maybe it's using something called Lun masking. And if it's configured that way, it's not going to present any limitations to this particular Six host. And so let's take a look at the background here. I have actually set up my Windows server to be an Iscz server. So if I go to server managers on this server2016 instance and I go under Is Gaze, we can see here that there are a couple of volumes that I'm actually sharing out as Is Guzzy. So why is my new Six host not detecting these volumes? So what I've basically done here is I've grabbed a couple of drives and I've made them lunge. I've made them scuzzy and available. And I'm just going to go to this one particular loan here. And you can see here that I've configured this one particular Lon with certain access servers. I'm making it available to certain Izzy initiators and my new Six host isn't listed there. So because of that, I'm not allowing that host to communicate with this one.

And so that's my problem from an iSCSI perspective. That's why nothing is showing up. And if I scroll down here, you can see my eye-catching target. I'm going to modify that target. I'm going to go to properties. I'm going to go to Initiators. I'm going to add a new initiator and it should show me the initiators that I've attempted to communicate with this loan, and one of them is my new Six host. So I'm going to go ahead and select that host applies here and I'm going to allow that host to communicate with this particular lung. So now if I go back to the VSphere client and I go back to this host and I rescan my storage, it should now start discovering some of the lungs that are available on that storage array, and I'll wait for my rescan to complete. It's rescanned HBAs, it's rescannedVMFS, and look at that. Now I'm seeing lungs available to this Xi Host from that storage system. If I look at the data stores that are associated with this Six host a minute ago, I didn't have any. This is Gusy and this large shared datastore that is also available on all of the other ESXihosts that are connected to this is the Cosby storage system. So it looks like this operation has been a success. I've configured a new software isguzzy initiator. I've configured it for dynamic discovery and, as such, it reached out to this Guzzy storage system, discovered all of the available VMFS data stores that had already been created on that storage array, and made all of those accessible to this Es Xi host.

7. Demo: Create a VMFS datastore

So I'm going to click on my storage tab, and under my storage tab you can see I've already created a couple of data stores here. I'm going to right click on my training virtual data center. I'll go down to storage and I'll choose New Data Store. I'm going to create a new VM FS data store. So hopefully I have a loan available that I can utilize to create this new VMFS data store. I'll choose VMFS. I'll hit next, and I'm going to name this data store Discussion. So this is my first data store that I'm creating that's going to be a VMFS data store. And then I need to choose a host here. So I've only got one host in my environment at the moment. I'm about to add another one, but I only have one host at the moment, so I'll just choose that host. And here you can see all of the lungs that this host has available to it. Consider these lungs to be essentially these chunks of disc space, chunks of available space that are presented either locally or through some sort of storage array. And that's what we've got here, an iSCSI storage device that I can go ahead and create a data store on. So what I'm doing is I'm going to grab10 and the 9.9 gigs of capacity that it's offering me and I'm going to format that capacity and create a VMFS data store on it. All of the hosts in my lab have Six version seven, so they're going to work just fine with VMFS six. So I'm going to go ahead and choose VMFS Six because it supports automatic space reclamation, so it's going to be more space efficient. So Vmfsix is the latest and greatest version of VMware's file system that was introduced in VSphere six-five, and it has a few substantial enhancements versus VMFS five. But the most important thing that I want you to understand is that there's no direct migration or upgrade path. So, if I create a VMFS Five data store, I can't later upgrade it to VMFS Six. I have to just basically migrate all the virtual machines off of it to a VMFS six data store. So there's no upgrade path from VMFS five to VMFS six. Basically, I just create VMFS six datastores and then move everything to them.

And these VMFS six data stores feature automatic space reclamation, which is a critical feature, really important and is going to greatly reduce the amount of deadspace sitting on these data stores. And now, based on what you see on the screen here, it can be a little misleading. I just want to point out that both VMFS Five and VMFS Six support loans that are greater than two terabytes. So don't think that because you're choosing VMFS Six, you're not able to doa lawn that's larger than two terabytes. That's not the case. VMFS six also supports storage devices that are greater than two terabytes. And if you're looking to kind of dig into the differences between the versions of VMFS data stores,you can find all sorts of information on VMFSSix versus VMFS Five here at Docs vmware.com.Support for larger than two terabytes is available on both VMFS Five and VMFS Six. So that's the critical thing to understand here is that if you have the option and all your hosts are later than ESXi six, dot five, go with the VMFS six data store. And you'll notice that if you're doing ESXisix or earlier, VMFsix is not supported. So that's really the only case in which I should select VMFS Five at this point is if I still have some of those legacy ESXi hosts. Okay, so in this case, I'm choosing VMFS. Six. I am going to take all of the space available on this machine and dedicate it to this data store.

And here I can choose the block size and the space reclamation granularity. And so basically, what this is going to tell me is, hey, any deleted or unmapped blocks are going to be reclaimed on the next one at low priority. So basically, what this means is that space that is no longer being used will eventually get reclaimed from that loan and freed back up so it can be used again and again. This is a significant enhancement versus VMFS Five. So I'm going to hit next here and that's it. It's ready to be completed. So I'll finish this process, and my new data store should show up right here. There's my eyeskui. data store. It's been formatted with VMFS Six, and I can see the hosts that have this data store available right here. And at the moment, there's only one ESXi host that can access this data store. So I'm going to pause my recording for a moment and go ahead and add a second ESXi host to my inventory. Okay, so now I've got a second ESXi host added to my inventory. So I'm going to go ahead and grab thishost and I'm going to do this extremely quickly. I'm going to add a new software Isczzy adapter. I'm going to go ahead and set it up to query the same Iscz storage array that the other ESXi host is querying, just like we saw in the previous video. And then I'm going to run a quick scan of storage so that my new host can find all of the available ones on that storage array.

Okay, and there we go. After rescan on the second host,I'm now seeing these ones available. So here's where it gets interesting. I created a data store, and if you look at this host, 192, 168, 199, here's the Iscuzzy data store that I created. And then, after the fact, after that was already done, I added a new host to the inventory. And here we see on the second host that the data store is already available. I didn't do anything to add that data store to that second host. All I did was create an anIscasi initiator on the second host. I pointed that storage adapter towards the storage array that I wanted it to discover. And when it went through dynamic discovery, it discovered the available lunge on that storage array, and it found that on one of those LUNs, we've got an Icecazzy data store already created. So it appears that the Icecussy data store made it available on this host as well. So now I have two ESXi hosts that both have access to shared storage. They both have access to this Discuss data store one. And if I go ahead and add a new datastore, I'll just do this under hosts 192, 168, and 199. Eleven. If I go ahead and add a new data store, I'm going to grab another one of those LUNs off of my VMFS storage array. I'll call it cozy. DS Two. Next, I'll choose VMFS Six. I'll use all of the available space, and I'll hit finish. And now I've got a cozydata store two available on host Eleven. If I go down to twelve, it's automatically there as well. So that's the beauty of shared storage. When you create a data store that's available on one host, that data store becomes available on that other host, as long as that other host is also connected to the same storage array and has access to that same set of months. Okay, so now we've created a couple of eye-sucking data stores. Let's take a look at what else we can do with these.

So I can right-click an Icey data store and I can choose to unmount it. So let's do that. Let's unmount. This is it? DS Two. I'm going to pick one particular host to unmount from, and once that data store is unmounted, it will be unavailable on this second host. The other thing that you can do with the data store is So I'm going to right-click this data store one more time and choose Delete data store. Now, here's what's going to happen when you delete a data store. When I created this VMFS data store, what happened was a loan on my Discos storage device was selected. That lung was formatted with VMFS Six, and the data store was created. If I choose to delete this data store, what's going to happen is that all of the data in that data store is going to be destroyed. Everything in this data store is going to be permanently erased. It's going to actually remove the VMFS data store from that lung and destroy all of the data on it. So keep that in mind when using I scuzzy or any VMFS data store. If you delete the data store, not only are you deleting it for this host, you're also deleting it from this host, and you're actually deleting all of the data on the storage device itself. As a result, deleting a VMFS data store is irreversible. Once it's done, You've destroyed all of that data. If you had a backup, you could restore it from the backup. But deleting a data store is permanent, and it's going to destroy all of the data on that data store for every host that uses that data store. So just exercise caution whenever you have to delete a VMFS data store.

8. Demo: Storage Port Bindings and Multipathing Policies

In this video, I'll show you some storage configuration tasks that you may need to perform on your VMFS data stores. So in our environment, I'm going to go to hosts and clusters here. And here's my first ESXi host. I'm going to click on Configure. In an earlier video, we created an Iscuzzysoftware adapter, which we can see right here. And on the second host, we did the same thing. We created an Iscoszy software adapter and we configured dynamic discovery to allow it to discover the ones that are available on this storage array. So this is our target. This is the storage array, the storage adapter of the ESXi host that's also called an initiator. So now I've got this target identified and we discovered a group of LUNs shown here under devices that were reachable on that storage array. Let's click on the network port binding tab.

And on the network port binding tab, you can see I haven't configured anything here. So let's take a moment to think about the overall physical architecture of what we're working with here. So here's a diagram of our Iscz storage network. And on the left, we've got our ESXi host. So on this ESXi host, I've created a software. Is Gaze an initiator? That's my storage adapter. And at the moment, I've got one VM kernel port on a virtual switch. So here's my VM kernel port, and all the traffic for storage is flowing out of the storage adapter through this VM kernel port, through a single physical adapter, and out onto the ethernet network into this one host. And the one host has reachability to the storage processors of my iSCSI storage array. In a real-world scenario, I might want to identify a second VM kernel port and bind that second VM kernel port to that software icecuzzy. In that way, it can send traffic through the physical adapters of a second virtual switch. And we have a much more resilient design here that can tolerate more failures. So there are really a couple of ways that I can accomplish this. So let's focus on this method for the moment.

I now have two different VM kernel ports. I'll call this one VMK two. And here's VMK One on two different virtual switches with two different sets of physical adapters connected to two different ethernet switches. And what's going to happen is my multipathingplug in is going to handle distributing that traffic across those two VM kernel ports. So let's say, for example, I'm using round robin. What's going to happen is some of the traffic is going to go to adapter one and some of it's going to go to adapter two. And vice versa. It's just going to basically rotate between those physical adapters. So the storage adapter will actually do load balancing across these two different VMkernel ports based on the Multipathing plugin. So this is a really great design. This is the ideal design for Icecussy because not only does it provide you with redundancy,but it's actually using the storage multi-pathingplug in to achieve that redundancy. Now, there is another option here. If I wanted to, I could say I'm just going to forget about having two VM kernel ports. So technically, I don't really need that.

So I'm not going to bother putting an aVM kernel port on this other virtual switch. I've got my one VM kernel port here, and I'll just configure multiple physical adapters on this virtual switch, connect them to different ethernet switches, and then I've achieved physical redundancy that way as well. That actually works too. That's another option that we can do. So now my storage adapter is just pushing all of the traffic to this VM colon port. And then I've chosen a nick teaming method like load base teaming or IP hash to actually determine how that traffic flows over those physical adapters. Now, the reason why I typically don't use this approach is that I think about the effectiveness of load balancing. Here I've got one VM kernel port with one source IP. If I use IP hash load balancing, the traffic is often going to the same destination IP. Therefore, there is a very good chance that the vast majority of the traffic will end up going over a single physical adapter, whereas the other adapter will not be as heavily used. So that's why I kind of lead with the other approach. Let's create VM kernel ports on both of these virtual switches. Let's let the multi-pathing plug-in handle loadbalancing that traffic across those VM kernel ports. Okay? So now that I've talked about the theory behind this a little bit, let's take a look at how this works. So here's my software, Icecus, the initiator.

I'm going to go under network port bindings, click add,and what I'm essentially doing here is choosing a VMkernel port for this storage traffic to flow over. So the most important thing is that this VM kernel port is on a virtual switch that is connected to the physical switch that my Ice Gusy store is connected to. I don't want to push this traffic overlayer three times if I don't have to. I want to keep it on a controlled and as small a network as possible. So, ideally, my ESXi host has an aVM kernel port on a virtual switch. That virtual switch has a physical adapter connected to a physical switch, and that physical switch is connected to the isGuy storage array. And then ideally, what I would do is go ahead and install my ESXi host here. I would create an additional VM kernelport by clicking on Ad Networking. I would put it on a different virtual switch with a different physical adapter connected to a different physical switch, which is also connected to that Isguzzi storage. That way, I have redundancy and I have load balancing. So if we go back to our storage adapter, That's what the network port binding area is all about. And if we go up here to the data stores here, you can see all the data stores that we've got created. I'm going to focus on one of these: Scuzzy data stores. And so here's my first I Scuzzy data store. I can click on Configure and I can make some choices in regards to this particular data store. So I'm going to click on Connectivity and Multipathinghere and you can see that this data store is currently mounted to this ESXi host here. I actually unmounted it from Host Eleven.

And if we take a look at the area below it, you can see that this data store is created on it. Here is the path selection policy. This is that multi-pathing plugin that I was referring to. So at the moment, this particular data store is being accessed using the fixed path selection policy, which means a single adapter is being used. If I want to, I can go over here to edit multipathing for this particular data store and I can choose the appropriate path selection policy. So, if I have multiple adapters that I want to spread this traffic across, I can choose round robin. If I want to send all of the traffic across a single storage adapter and just use whichever one comes up first, I can use the most recently used. Or if I want to specify a specific storage path that should be used, I can use Fixed. So these are the different path selection policies that I can choose from here. And these are all built right into ESXi. These multipathing plugins are baked right into ESXi. So the primary recommendation that I have for you is to talk to your storage vendor to review their documentation and figure out what kind of multipathing policy is appropriate for your storage array. Some storage vendors have their own multi-passing policies that we can then plug in here and utilize. So the main recommendation here is to consult your storage vendor's documentation and make sure you've chosen the right multi-pathing policy.

9. Demo: Expand a VMFS Datastore

And I'm going to choose my Iscuzzy DS three. This is the data store that I want to make larger. So at the moment, if I click on the summary screen, I can see that my capacity for this datastore is 9.75GB and I've got free space of 8.34GB. So what I want to do is make this thing bigger. So I'm going to start by going to my Iscui storage system and expanding the Lun. The loan itself has to be bigger in order for me to expand this data store because, at the moment, the data store is using all of the available space on that loan. So what I'm doing in my lab environment is actually using a server 2016 virtual machine for my domaincontroller, and I'm also using it as my Iscuzzy server.

This is my Icecazzy target where all of my Ice cuz I data stores are hosted. So at the moment, you can see I'm shutting this virtual machine down. What I'm going to do is I'm going to expand the physical discs of this virtual machine. And what I'm basically doing here is simulating. Hey, I just put more discs into my eyescu storage array. I've just expanded the physical capacity of my Isczy storage array. I'm basically mimicking that by adding more storage to this server virtual machine. And then I'm going to use it to make my Lun on this Iscuzzy server larger. Okay, so now my Server 2016 VM that I'm using as my I Scuzzy system is done being shut down,I'm going to go ahead and grab one of these hard discs and I'm going to expand it. So it's currently ten gigs. I'm going to make it 20, and I'm going to increase the size of this virtual disk. And what I'm doing is basically the virtual equivalent of adding physical storage.

So if this was an Iskuzzy storage array,what I would have basically just done, what I've just mimicked, is add more physical capacity to that Icescuzzy storage array. So now give us a few moments to boot up here. And when it's done, what I'll do is I'll go back to my Vsphere client and I'll rescan my storage so that it can detect this larger lung that is now available. Alright, so now that my Windows machine is finished booting up,you can see that one of my discs actually has 20 gigs of capacity, ten gigs of which is unallocated. So what I'm basically doing is the equivalent of making my lun larger. I'm now going to allow this volume to leverage all 20 gigs of that space. So now that I've increased the size of my Lun,what I did next was I came in here to my ESXi host, I clicked on configure, I clicked on storage adapters, and I performed a rescan. I rescanned all my storage adapters on this host, even though I really only needed to rescan this one. And then I refreshed my capacity under the Devices tab for this adapter. And you'll notice it now reflects 19.9 gigs of capacity. So that lun is now larger than it was before. So IscosyDatastore Two is actually not using the entire loan. Let's click on IscosyDatastore Two and notice the storage capacity for the datastore itself has not actually changed. So the LUN that this data store exists on is bigger, but the data store itself is not bigger.

So let's click on Configure and under General for this data store, I'm going to click to increase capacity, and it's going to ask me which one I want to use to increase the size of this data store. It's the same one that it's already using, but there's an extra 9.9 gigs of capacity available there. So I'm going to use the extra free space that's available now on this one to increase the size of this data store. It's currently at 9898 gigs. I'm going to add ten more gigs to the thisdata store, and I'll click next and finish. And there we go. I've expanded the size of my VMFS data store by increasing the size of the underlying loan and then going into the Visa client and increasing the size of the data store there. I can expand this data store and make it bigger. Now, there are other ways that I could make this data store bigger as well. Like, for example, if I had more LUNs available,I could hit increase here and select a different one, and I could create a data store that would actually span multiple lungs under most circumstances. We really don't recommend that you have a one-to-one relationship between your data stores and your loans. So ideally, that's what we want to configure, one data store per loan, and maybe maintain that one-to-one relationship between them.

10. Dependent Hardware iSCSI and Six 7

In this video I'll explain a dependent hardware I scszyinitiator and how it can be used by the Sexists to communicate with anSic storage array. So here we see in our diagram of virtual machines on the left with a virtual SCSI controller. As that virtual machine issue uses storage commands, they flow through the virtual system controller and hit a storage adapter that prepares them for transmission over a network. In this case, we have a physical storage adapter. This is different than software, I scuzzy. Where the storage adapter is based on software. Here we've purchased some actual hardware that's going to handle that workload. Now, with dependent hardware, is a cozy initiator. It's not a complete hardware solution, it’s a physical storage adapter. However, it still needs to be bound to the VM kernel ports that are created in software. So while the dependent hardware is scuzzy, the initiator eliminates some of the CPU overhead that's present with software is a scuzzy initiator, it doesn't eliminate all the CPU overhead from the Six host entirely. To do that, we would need an independent hardware Ice Gazzi initiator that includes its own physical Ethernet adapters.

Prepared by Top Experts, the top IT Trainers ensure that when it comes to your IT exam prep and you can count on ExamSnap Professional VMware vSphere 7.x certification video training course that goes in line with the corresponding VMware 2V0-21.20 exam dumps, study guide, and practice test questions & answers.