Red Hat OpenShift Red Hat OpenShift 4, 3 Apr 2020

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Using Local Disks on VMware with OpenShift Container Storage

Description

Chris Blum, Senior Architect in Red Hat Storage, demonstrates using local disks on VMware with OpenShift Container Storage 4 in OpenShift 4.

For more information, please visit openshift.com

A

Hi, my name is Chris bloom and I'm, a senior architect in storage per year. Today, I'm going to talk about using local disk on VMware with OCS 4 on VMware. We have multiple choices of using local disks for one we can use regular vm decays, which are being this that are directly attached to VMs.

A

They can be based on pretty much anything. Then we have the choice of using our DM or raw device mapping which Maps certain rate devices directly into a VM, and we can also use VM direct path, IO, which can map device local devices to directly to VMs, and this is actually what we're going to look at today. So we will try to map local and many devices into a VM, and this is probably one of the hardest steps that will involve many things to do.

A

If you just want to use being the case, then you just do not do the first couple of steps in the beginning, so there are some requirements that you need to satisfy to do that. Yeah install the ocean, Explorer operator, you install a local storage operator.

A

You need to have local disks of some kind and at least three hypervisors, like I said we will be using local and B media voices for the jewelry path I/o, and you also need the ability grade, three ten gig being the case on the same hypervisors that you dedicate for OCS for the sec monitors.

A

So how do you set up the local disk? At CVS? You assign the OCS label to the notes with the local storage install the local storage operator in it's namespace you, you can use the daemon set to figure out the disk IDs for the local disk under the unity case, eater yeah, by using a demon set.

A

That's provided at this link here or you can do that yourself, either by SSH against the host or using OCD buck node, and then you create two local volumes and local search operators namespace one with the volume hood filesystem for the set monitors and one with the volume mode block for the local disk and finally, to use these local disk Vivi's with OCS. You create a storage cluster. So this is the full demo that you need to create.

A

I've split it up into two sections, so basically, you put in here the storage class name for your local PBS, with a file system mode, and over here you put your source class name for the PBS, with the volume would block and that's basically it so now, let's look at how this actually works with a little demo.

A

So first I have prepared this open shift cluster. Here, it's freshly installed with OC open share for 35 on VMware, and this is the VMware cluster, so we're on per seven. These are all the notes that make up the cluster. We have three masters and six compute nodes. We will be using the first three compute nodes today to attach the local disk, especially the nd case, especially the VM direct pass AO. We need to shut down these workers because I know that nothing is working on the cluster.

A

I can just tell to shut down if you would be doing this on a running cluster than obviously first drain those notes and ensure that the pots are properly migrated off of these notes before shutting them down. So when they're shut down, we ensure that these are running on the three hypervisors, so you see computing. No zero is running on host raipur three, so I want to ensure that they are all running on different hypervisors.

A

So you see that compute one is running on to compute. Two is running on three, so I'm going to move compute zero over to the first hypervisor each of these hypervisors one and minidisc that I'm going to touch, and now they should all be on different hypervisors. So this is fine. Now, all right, so let's attach our local nvme for this I'm going to add an nvme controller and I'm going to add a PCI device- and you see over here.

A

This is my PCI device and I need to reserve all memory for VMware to forward that PCI device. So I'm going to do that and then I'm also going to add the 10 gigabyte VMDK for the sat monitor.

A

A

And I'm going to do the same thing for the other two workers.

A

Also note that when you're attaching a PCI device, that's passed through PCI device, then obviously there are some limitations on that VM. So not only do we preserve all memory and hard allocate that to the VM, but it can also not suspend migrate with motion or take a snapshot of that virtual machine, but that shouldn't be too bad for you.

A

All right now that we're done this, we power them back on.

A

And they should eventually be read again and a cluster. You can either monitor their stay in the web UI or you can do that from the terminal when you're logged in. If you don't see that your nvme PCI devices in the VM options, then you probably haven't configured them for forwarding to forward them. They need to be.

A

Configured for pass-throughs.

A

You do that when you select one of your hypervisors, you go to configure under Hardware PCI devices. Your nvme device needs to be listed here as pass-through enabled device. You can configure PCI devices for password by clicking this button up here and then selecting your nvme device like this.

A

Do note that when you have your nvme configured for pass-through, it cannot serve as a data store anymore. So it's going to be dedicated just for that single use case.

A

So, let's hop over to the terminal briefly.

A

So I'm already locked into the cluster, so I can list the nodes and we can see that the workers are already back online and ready for new workloads and for this demo we have four files. So first, let's look at the requirements.

A

We first create the namespaces for the two operators that we are going to install and then we are all actually going to install them. So let's do that.

A

So obviously the avengers' storage namespace is for the OCS operator and the local storage namespace is for the local storage operator and then we label the opens of storage namespace for monitoring so that permit us knows it should be monitoring. Let's hop back over to our web UI over. Here we go to operators operator, heart.

A

We go to storage, we select a container storage, install that we install it to our open storage namespace and we leave automatic.

A

And then we go while this is installing in a background, we can go back to operator heart.

A

We go to storage and now install the local storage operator.

A

And this will take a second.

A

All right and now that this is installed, we can go back to our terminal.

A

And in here we do have two files, the local file and local black llamo, and we need to feed it the device path and, as a best practice we're using by ide device path, because the disk IDs could change between reboots and so that we can easily find out the disk ideas and all the nodes we can use to be to do Dax. We first need to label our notes so.

A

We get the note names and we want to. We know that the first three working notes had the local disc, though we will just do OC.

A

Label note compute 0 and the OCS label.

A

And we do that for the first three workers.

A

This is important so that the demo set nurse on to which notes to look for a local disc. Now that we have that we can verify that the low label worst successfully applied by just asking for the notes with the labels. So we see the first workers have the label. This is fun so now we deployed in human set. The demon sir looks like this looks like this, and you see it's nothing too fancy it's just running some bash to figure out the disc by IDs, going back to the terminal.

A

Let's apply this.

A

The link to the demon set jung-hwa is in the presentation, slides, so let's check out the parts, and we see immediately that parts burn spin up on nodes.

A

Since we have three workers with label, we have eventually three three of those steam and certain parts and the running are this is fine. So now we can go and gather the lock output of those pots.

A

And this looks like this, so we have three workers compute one computer compute two and they each have three disks. So I know that SDA is my root disk, so I'm not going to use that s. Db is the lien BK disk and nvme zero and one is the enemy disk and they are fortunately named the same on each note. So what I can do is I can now use these disk ideas in my files.

A

So let's do that I'm just going to copy all this output and then we're using them to go to their local file.

A

Delete this and I know I just pasted here.

A

And I can basically remove all the lines that I don't need. So this is for the Vemma case, so I'm just going to remove the SDA and the NV meet and.

A

Then obviously, after uncommon, this we can leave the comments if you want to, but they will be ignored by the OSI command. So this is fine. Now, let's head over to the.

A

Local black enamel and we do to a very same thing.

A

So this time, because we want the enemies which is going to remove s da and s DB.

A

And comment this fine.

A

Let's have a last look at our files.

A

So that local volume with the volume load file system has the SD beetus and they're mentioned by ID and local volume, where the volume mode block has the device pass for the enemies and for all three nodes, very good. So let's apply this.

A

A

So what does yes? We should now eventually see. There is.

A

So now we see the local PD is slowly appearing, so the 10 gigabyte capacity TVs are used are based on the VMB case and the 1 point. 5 terabytes TVs are based on our local enemies and, as you can see, that the 10 gigabytes TVs are using the local file, storage class and nvme devices are using the local block storage class, just as we want it, and we are going to use the storage classes now to deploy or see us on top of these devices. So this is going to look like.

A

This is going to look like this, so we are going to apply an OCS storage cluster and we're again going to tell OCS to use the local file storage class with the volume mode filesystem for the monitors and we're going to tell OCS to use the local block Sturge class with the volume earth block for the data pots. The parts with the Ceph OSD is done and we want three of them.

A

Obviously, these devices are not portable, so we tell us: yes, do not try to move parts between nodes.

A

So, let's first look at the parts and the open chef, storage, namespace, they're, all running the soloist on and now we apply this file.

A

And what we're going to see is that this will spawn pots and create TV seeds that match those and weenies.

A

A

Now that all pods are in state running, the OCS deployment is done and let's look at this again outside a watch. We have quite a bunch of parts, the point now and we have the OS DS and we see down here that the OS DS are now also using the enemies. We see that for the capacity we can match that PV back to our local disk and they're, using our local clock through squads and for a monitors we're using the local file storage class.

A

That gives us 10 gigs the indicator directly attached to the VMS and obviously OCS is the point now and we do have access to the osseous storage classes here and also the new bio BC storage class for OB ceased and just like always, we can now go back to the web UI on to the dashboard, and you see that everything still looks healthy. We can go to the newest, yes dashboard here, and it behaves just like and OCS the point with the thin storage class on vmware, and we have the very same look here.

A

This is probably just unhealthy right now, because it's rebuilding your data. Resiliency was just a ploy but will normalize in a bit, and this concludes the demo about local storage on lemur. Thanks for watching.