Skip to playerSkip to main content
  • 4 days ago
Transcript
00:00Welcome to this demo on OCI block volume service. In this particular demo, we are going to attach
00:11a block volume to a running instance. And then what we are going to do is we are going to size
00:17up that volume to a bigger size using what is referred to as online resize. So let's get started.
00:23I am logged on to my OCI console and to bring up the block volume service, I'll click on storage
00:31and under storage, you see block storage, object storage and file storage listed. So click on block
00:37storage and this will bring up the menu to create block volume. So you can see here create block
00:43volume is there. The button is there and I have no block volumes running right now. So I'll click
00:48on create block volume. And it asks for a name. So I'll say this is block volume for a demo. I can
00:55create this in the sandbox compartment. That's the compartment we have been using for all of our
00:59resources. And it picks an availability domain. It's an availability domain specific service because
01:05you are getting a storage remote storage disk in a data center. So that's what is being shown here,
01:13the different availability domain. So I'll pick 81. And right here, I can go with a default size of
01:191024 gig or I can go with a custom size. So let me pick up custom and let's say I want a disk which is
01:26only 100 gig in size and I can go all the way from 50 gig to 32 terabytes. And below that, you see these
01:32two toggle switches which are on auto tuning. So what are these? Let me just explain very briefly.
01:40So in case of block volume, there is something which is called block volume performance unit.
01:45And basically, this includes the concept of volume performance unit, what is referred to as VPUs,
01:51what you see here. And the idea here is you can purchase more VPUs to allocate more resources to
01:57a volume, increasing your IOPS per gig and throughput per gig. So as you can see here,
02:03if I kind of, first thing you see here is if I'm getting 100 gig and if I scroll down here,
02:11there are the various kind of a slider. So if I go with a lower cost, I can, you can see that I get
02:17two IOPS per gig. That's the lowest tier. If I go to the next higher tier, if I just move this
02:24mouse here, you can see this is the balanced tier. And balanced tier basically gives me 60 IOPS per gig.
02:32So for 100 gig volume, I get 6,000 IOPS. And if I go to the next tier, which is higher performance,
02:39I get 75 IOPS per gig, which is, which is for 100 gig volume, that is 7,500 IOPS. And as I start
02:49moving this slider, there's something called ultra high performance. And for ultra high performance,
02:54I go all the way from 90 IOPS per gig. So you can see that here, all the way to 225 IOPS per gig. So
03:03for 100 gig volume, I can get a maximum of 225,000 IOPS. And as I increase the size here, let's see if I go
03:12from 100 gig volume to a 2000 gig volume, you can see I hit the maximum threshold, which is 3000 300,000
03:21IOPS, which is just amazing. It's a lot of IOPS for a volume. So I can do that. Let me switch it back to
03:27100. And then there are two auto tune capabilities here. Basically, the idea is if a volume is throttled
03:34and doesn't get the desired performance at its current perform performance setting block volume
03:40or auto tuning here, actually, let me just turn it on basically gives you gradually increases and
03:46continues to monitor the volumes performance in increments of 10 VPUs per gig up to the maximum
03:52VPU. So I can specify all the way from 10. Let's say I can specify to 40. And what this will do is it
03:58will set the thresholds from 10 to 40. And it will let me handle spikes, which need, let's say, you
04:04know, 9,000 IOPS or 10,000 IOPS. So I could actually do that. And so that's basically the
04:09performance based auto tune. And there is also something called detached volume auto tune, which
04:14as you can read here, turning on detached volume auto tune will change the volume's performance
04:19automatically to lower cost when it is detached. When it is reattached, the performance is
04:25automatically adjusted. So these are just features which help you save costs. So I'll leave these as
04:30is 100 gig size is fine. And on here, you can set things like backup policies, cross region
04:36replication, whether block volume uses Oracle manage keys to encrypt the data, which is always on by
04:43default, or you want to use your own vault and your own master encryption keys to encrypt the data.
04:50So I'll leave the default on here. And I'll click on create block volume. And you will see that
04:54within a few seconds, the block volume will be created. And as soon as it is created, I'll go
05:00ahead and attach it to a compute instance. If you recall from the previous demo, we had an instance
05:06created in this region in availability domain one. And if I access this, I'm connected to this instance
05:14using SSH SSH. And if I if I list all my block devices for this particular instance, you can see right
05:23now I only have the boot volume, which is 46.6 gig. And this is the the boot volume is a kind of a block
05:31volume where operating system is kept. So when the instance instantiates stands up, it needs that operating
05:39system. And that's also kept on a remote storage disk referred to as the boot volume. So that's all which is which is
05:46there. But once we attach the block volume to the instance, you will see that I can see that volume listed here as
05:53well. So let me just refresh this page and see if the block volume is up and running. And you can see it's it's available.
05:59So if I click here, now, it's pretty straightforward to attach it to an instance, I can see all the
06:04characteristics. And if I click on attached instance here, you can see there is no instance with to which
06:09it's attached to I can change that click attached to instance. And I now get two attachment types iSCSI
06:17and para virtualize iSCSI basically means that we use the internal storage stack in the guest operating
06:24system, not the hypervisor for attachment purposes. And para virtualize basically means that virtualize
06:30the volume in hypervisor and attach to the to the instance as you can read here. So we'll choose a
06:35para virtualization. And then access type could be read write or shareable. Now we're not sharing this
06:41block volume with other instances. So I'm not using any of those options. Read write is fine. And then I can
06:48choose an instance here. Now remember, this instance is running in 81. And that's why I was able to choose
06:54that this particular instance. And then we also need to choose a device path. So that it helps with
07:00instance reboots, I can use a consistent device path when the instance reboots, etc. So let let this get
07:10attached. And as it's, it's it's doing so next steps for me would be to go ahead and partition the disk,
07:18I could format the disk, install an operating system. And then I can I can mount that disk
07:26to the instance. But this is a foundational demo. So I'm going to skip those those steps. If I refresh
07:33the page, let me see if it's installed. Instances, it's still in the it's now you can see it's attached
07:41here. So if I go back to my SSH window, and if I run this command, again, you will now see that the
07:50100 gig disk is available. And I could create a mount point as right now it's missing. But but you
07:55get the point that the disk is now available. And I can use it to store my data and and perform all kinds
08:05of operations. This was a quick demo on how to use block volumes. I hope you found this demo useful.
08:11Thanks for watching.
Be the first to comment
Add your comment

Recommended