00:10
Welcome back to Digital Bytes. I'm your host, Ralph Ronzio, the technical director of the Digital Experience team here at 色控传媒. I'm super excited about this episode because it's another demo day. We're gonna look at a new feature called workload placement recommendations. Now this is sort of an evolution of the planning tool.
00:26
So up until now, the planning tool can scale workloads. It can show you what would happen if you migrated workloads, if you clone workloads, and you can even bring on 13 predetermined workloads based on workload DNA. But what this does is it tells you where you should put it. So if you have a a fleet of systems,
00:43
it's gonna tell you the best system for the job. This is really, really cool and it's taking the planning tool to the next level. So let's dive in and check it out. OK, so. logged into Pier One, and when you first log in, you're going to see this banner and if you scroll through this banner, you're gonna see a bunch of the new features of which workload
01:00
placement recommendations is one of them. So you could read more about it, which would open up another window which gives you a little bit of uh information about the feature and then you can look at the release notes. So what we're gonna do is we're gonna dive right into the tool itself. So under analytics you go to planning.
01:14
The scenario we're gonna run right now is we're gonna choose an array here from our list of arrays and we're gonna go to simulate workload. Now what I'm gonna do is I'm gonna simulate that we want to migrate. The workloads running on this system to another one somewhere else in our environment. Now that's not an easy task because you have to take into consideration a lot of things and if
01:34
you have other systems in your fleet, it's just tough to evaluate yourself. So when we click on the migrate button, you're gonna see it says recommendations across 32 arrays. So I have 32 arrays in my fleet, so I can hit show recommendations and what the engine is gonna do is it's gonna run now and it's gonna find me the optimal placement.
01:52
Now there's several icons it's gonna display. There's an icon that's green and then there's uh that means it's the optimal placement, so that means this array is one of the best targets for what you're trying to do. There's just a straight line and that's gonna be non-optimal and then there's a circle with an X in it and that's actually it doesn't fit on any of the arrays in your environment.
02:13
So let's wait for this to complete and see what array is most optimal for this workload. OK, so it came back and you can see the top 5 here are actually the most optimal for that workload. Now if you scroll down, you'll see the line means non-optimal, and then the circle with the X means it doesn't fit at all. So non-optimal actually means the array can handle the change in capacity,
02:37
but there are potential better options than that particular array. So it shows me there's 5 really great options. So what I can do now is I can select a target and then hit next. And now it shows me the before and after for that target array. So it's great. It, it shows me exactly what will happen.
02:54
So these are, these, these are great scenarios for existing workloads, but what if you want to bring on net new workloads? So I'm gonna cancel out of this. And I'm gonna go to simulate new workload. So let's say that I wanted to add, let's see, one of these predetermined workloads in here,
03:10
we'll choose, we'll choose postgrad SQL. So I'm gonna choose that and I'm gonna choose a large deployment. Now again, these sliders can be adjusted. We covered that in a prior episode of Digitalytes. You can go back and see that if if you're interested in understanding how that all works. You can click save.
03:27
And now what we're gonna do is we're gonna say, here's my postgraph SQL workload. Let's select the target array, but I don't know what one I want to to put the workload on, right? I want the system to evaluate it for me, so I'm gonna click on show recommendations again and it's gonna scan the entire fleet and make a recommendation for this new postgras SQL workload.
03:45
And again, so this means we can do it for workloads that are preexisting that we want to scale or migrate or brand new workloads that we maybe haven't included in our environment um until now. So let's give this a second to come back. And then we'll see what, what array will have the most optimal placement. OK, so only one system is optimal placement for that.
04:06
So I'm gonna select that one and as you can see there's a bunch that do not, uh, meet the meet the demands of that particular workload. Also for, for feedback purposes, there's this little thing that says how useful are these recommendations. I would definitely encourage you if for some reason you're not seeing something you would like, choose one of these and provide your
04:25
feedback that only helps us make the tool better. So then when I click next, it's gonna show me a simulation of those. Changes it shows me the projected load of my target array before and after, and I can save an exit. So currently right now it doesn't actually migrate the workload for you,
04:41
but it does tell you the best place for it. Now maybe in the future we'll do something like that. I hope you enjoyed this quick and easy demonstration of this. We will deep dive it again in the future, but thanks so much for checking it out with me today. Well, that about wraps up this episode of
04:56
Digital Bytes. Thank you so much for watching. I am so excited about this feature, and I think a lot of you will be as well. How neat is it that the system can now tell you where to place workloads and it completely takes the guesswork out of it. This is. What I'm talking about when I talk about AI ops at its finest,
05:13
it's this kind of stuff right here. So hopefully you're as excited as I am. Put your comments below. I want to hear from you. Consider subscribing to the 色控传媒 YouTube channel, and that's it for now. We will see you on the next one.
![Lightning bolt icon](/content/dam/purestorage/online-assets/icons/lightning.svg.imgo.svg"/){kind=icon}
