Cloud modeling capabilities in Foglight Evolve

Hi, and welcome to this really quick update on Foglight for Virtualization 8.8.5. My name's Chris Jones. I'm the product manager for Foglight for Virtualization. And today, I'm going to take you through a really short, two minute video on our cloud modeling capabilities. Let's jump in. On the screen, right now, we're looking at the cloud modeling feature within a capacity director. What it does is, allows you to model moving any number of VMs into any number of instances from any cloud provider or MSP. And let Foglight tell you exactly which VM should be in which instance type. And then, you can drill in and see the cost savings. Or maybe the cost increases from migrating to the cloud. So in this table, here, we've got 272 instances across Azure and AWS that are available in the US that we're going to model against. This data, we actually pulled live from Azure and AWS APIs. Over here, I've selected 142 virtual machines that are a part of this model. You can see them listed here. In these next three configurations, what I've selected to do is, down here, I am asking Foglight, put each VM in the best fit for itself. Above it, I'm saying use business hours data. Because I know my VMs are only busy during the day. And not during the night. If I was to use 24 hours worth of data, our peaks and troughs would be greatly diluted. Which really impacts our risk model, at the top. Which I'm saying, I'm not OK with any risk. So I want to spend as much money as it's going to cost to make sure each one of these VMs is never impacted by constrained resources. So selected my instance types, selected my VMs, I've set my risk level, chosen the hours for the data that we'll use in the model. And now, I'm going to say, find the best fit for all of these the VMs. So let's give it a go. So this model's running in the background. And it's taking all of the inputs that we've provided to it. 272 instance types, 142 VMs. And we're going through and looking at every single day, and every piece of data, within that day. It's the last 90 days. And when you use those 90 days worth of data to find the very best fit across all those tiers, across all those VMs, for each of those virtual machines. The things that don't fit, what we do is, we call it out. For things that do fit, we put into the model, that you can, then, drill into deeper and expose the true cost savings, or the cost increases, that will come from it. So this table, at the top, is what we can do. But just to show you, down here, very quickly, this is a list of VMs that we can't move. And it tells you why we can't do it. So in this example, it's, actually, saying that the OS that this VM is running isn't supported in that particular cloud instance type. So let's drill into one of these models. $2,000 a month to run 17 VMs. I think we should explore that a little bit further. With quick review model, we're going to see a collar at the top, here. We're seeing that we've got an overall fit for this model, for this instant type, 17 VMs. And we're saying it's poor. What that means is that overall resource utilization of that instance type, by each VM, is actually quite low. So it's less than 50%. What we can see is that most of the CPU is somewhere between 0% and 50% utilized, right? Which is fine. But really, what you want to make sure is that you've got a good fit on something. And that's what we do have from memory, here. We've got a really good fit, around 50% memory utilization in these green ones. And these blue lines represent someone, a virtual machine, that's using this greater than 75%-80% of the memory utilization. That's spot on. That's right where you want it. We can see here, that the IOPS for these VMs is, actually, quite low, as well. Now, this particular instance type could be memory performance churn. So it could be the cheapest one that's available. While we're there, what we'll to do is, come over here and see the on premise cost and the cloud cost per month. Now, depending on the OS type, you're going to see a fluctuation of these cloud costs. And the on premise costs, they're going to be based on what you've set up in cost director to model out your on premises costs. So we can see, here, that some of the cost more and some are, actually, going to save us money. To then take this to leadership and communicate it to someone, you could customize a quick export to something like a PDF. And all of a sudden, you can put this in an email and send it straight off to your team to say, look, I've run the model. Here's the best fit for these VMs in Azure. And should we start digging into this a little bit further and decide if we really want to do this model. And that's it. Thank you very much for your time today. And stay tuned for more videos on Foglight for Virtualization.