Search
Follow me:
Listen on:

Day Two Cloud 123: Managing Multi-Cloud Applications And Infrastructure With vRealize Operations Cloud (Sponsored)

Episode 123

Play episode

Welcome to Day Two Cloud, where the topic is visibility. Hybrid cloud visibility with a side of Kubernetes, to be specific. VMware has come alongside as today’s sponsor for a discussion about vRealize Operations Cloud to give you that visibility. “Visibility into what?” you might ask.

Visibility into your workloads and whether or not they are performing optimally. Are they over or under provisioned? Visibility into your dependencies, so you understand how applications are impacted by problems with various infrastructure components.

Visibility into your capacity, so you know if you’re gonna need more before next year, which is sort of a big deal considering global supply chain problems.

You get the idea. Our guests to help us get a handle on vRealize Operations Cloud are Sujatha David, Group Product Line Manager, Product Management; and David Pham, Senior Product Marketing Manager, both at VMware.

Show Links:

vRealize Operations Free Trial – VMware

vRealize Operations Features and Capabilities – VMware.com

Pathfinder.vmware.com – VMware Information Site

Transcript:

[Please note the transcript was generated by a program and has not been checked by a human for errors.] [00:00:04.550] – Ethan
Welcome to Day Two Cloud, where the topic today is visibility hybrid cloud visibility with a side of Kubernetes to be specific, VMware has come alongside as today’s sponsor for a discussion about Vrealized Operations allowed to give you that visibility visibility into what you might ask. That is a fair question. Visibility into your workloads and whether or not they Azure performing optimally. Are they over or under provisioned? Visibility into your dependencies so you understand how applications are impacted by problems with various infrastructure components visibility into your capacity.

[00:00:37.530] – Ethan
So you know, if you’re going to need more before next year, which is sort of a big deal considering global supply chain problems, you get the idea. We’re going to talk about it in more detail. Aks, we go our guests to help us get a handle on Vrealized Operations Cloud. Our Sue, Jonathan David, director of product management, and David Fan, senior product marketing manager, both at VMware. Sujatha and David, welcome to Day Two Cloud Suja. I want to direct the first question to you for those folks who maybe don’t know what Vrealize Operations and Vrealize Operations Cloud is all about.

[00:01:09.230] – Ethan
Give us that pitch. Give us the overview of the product.

[00:01:12.570] – Sujatha
All right. Thank you, Ethan. Thank you for having me and David on the show. We are super excited to be here first and foremost. And I love the introduction that you already gave about our product. So let me build on that. We realize Operations truly provides the self driving management to manage increasingly complex multi cloud environments. Now you’re talking about the complexity where customers are expanding their footprint beyond just their private cloud. They want public cloud. They are using their own managed clouds and so on and so forth.

[00:01:51.520] – Sujatha
So we realized Operations provides that consistency, that they can use one tool to manage these multiple different infrastructure gives them the way you said this was an earlier visibility to get that visibility of their workloads across these various clouds and give them automation, security, governance to properly manage their systems and those most important critical applications across multiple cloud environments. Okay, let me also talk about the personas now. Different personas may come and use the stool. They may be folks that are focused only on capacity management.

[00:02:39.190] – Sujatha
There are others who Azure very interested in the performance of the applications. The tool is built in such a way that any of these personas can come specify their business intent and operational intent and manage it that they want to manage that infrastructure. So let me pause. I can go on and on.

[00:03:02.610] – Ethan
I got a lot more questions to ask you here. You definitely hit for the marketing buzzword bingo cycle. So we got a lot of them in there. So during the rest of the show, we’re going to get you to defend the buzzwords and explain more nerdy technical detail what’s really going on here. But we get the big idea. We realize Operations is all about. Again, that visibility of workloads across a bunch of different cloud environments in one unified tool that can look differently depending on who’s looking at it and what sort of data that they need to consume.

[00:03:33.710] – Sujatha
Yeah.

[00:03:34.330] – Ned
My question that I want to jump in here with is I think Ethan mentioned at the top we’re also talking about do you realize Operations Cloud? So I’m assuming that brings something else to the offering beyond the things that you just talked about. What is berealize Operations Cloud versus just the V Realize software offering?

[00:03:53.680] – Sujatha
Absolutely. So the Vrealized Operations product has been in the market for more than eight years. It’s a pretty mature product. And hence, we have really great maturity in the offering. And the word that we use for we realize Operations product is the self driving operations aspect of it. There’s self remediation. There’s a whole bunch of technology that’s been built into the product. Now, the SaaS portion of it or the cloud is essentially to stand for providing all of the self driving operations as a service. Now, what does that mean for the user?

[00:04:37.870] – Sujatha
It means that you don’t have to worry about even installing the product, maintaining it, upgrading it and having that hardware that you need to set aside if you need to manage large workloads. So we Azure will come and do it for you. So essentially, you’re outsourcing the management of the tool that manages your workloads. And so Vrealize Operations Cloud is essentially software as a service of the On Prem offering, which is we realized Operations got it.

[00:05:15.320] – Ned
Okay. Very good.

[00:05:16.720] – David
And the awesome thing about that is you just get that faster time innovation. A lot of the customers are excited to get the latest features. We make these announcements like, oh, shoot. We have this schedule maintenance window. Do we have to stage everything and make sure that it’s not going to crash environment and spend time for after hours on the weekends? That sucks for a lot of the folks that have to manage to deploy these out. But again, we do it for you. So you get that flexibility looks awfully straightforward.

[00:05:44.460] – Ethan
David, to do the install for the SaaS flavor here for VR Operations Cloud V realize Operations Cloud. It was basically instantiated feeding some credentials, stand up a cloud proxy so that the SaaS instantiation can get to whatever the resources are it’s supposed to manage. And that was kind of if and then it starts gathering metrics and off you go. Well, let’s talk about the services that can be managed with vROps Cloud. Okay. We talked about you can manage workloads in a bunch of different places. Well, this is a VMware product.

[00:06:14.750] – Ethan
Does that mean I’m limited to kind of like the VMware world, Bsvp and BMC and VXLAN and that kind of stuff. Or give us the overview.

[00:06:23.730] – David
We support extended list of native and third party management PaaS so you can connect from everything from your physical devices, your Sans routers and switches all the way up to the application layer, whether you were in virtual machines, Kubernetes containers, and all the way up to the public cloud, the public native cloud, such as AWS. Microsoft is your Google Cloud platform AKS for additional insight and information. So, I mean, we’ll talk across the whole entire data center stack. It’s top to bottom.

[00:06:57.630] – Ethan
You said something important there that stuck out to me. I was thinking of this because I think in terms of infrastructure first, and it handles that. But you also said application, so it moves up the stack so I can throw, I don’t know, SQL Server or something at this. And if there’s a plugin for it, it will tell me SQL Server level kinds of things that are interesting to me.

[00:07:19.480] – David
Absolutely. So it’s an extensive list of applications and third party manufacturer that’s going to give you that complete visibility. And another thing is that through our recent acquisition of Blue Madora, we have what we call the Realized True Visibility suite. I mean, that’s going to give you all the extensive visibilities and extensibility, too. But I’m going to want to talk about complete infrastructure AKS, the infrastructure to the cloud.

[00:07:44.930] – Sujatha
Let me add a couple more points to what David just said. We know our customers use all kinds of application performance monitoring tools and operations. We allow you to have those integrations if you’re using Dynatras and Dynamics new rally, whatever your choice, what we let customers do is bring in that data into V realized operations. So again, as I mentioned earlier, our goal is to have give our customers one tool that they can use to manage all of their infrastructure and application. And so there is integrations to these other application monitoring tools.

[00:08:25.510] – Sujatha
And also with our recent enhancement to include open source Telegraph Agent, we now have support for 200 plus applications, so customers can come enable management packs, or if they have the open source Telegraph and it’s up to you, whatever you want to manage, bring those metrics into viral operations and manage it holistically from one place.

[00:08:54.640] – Ethan
Yeah. This gets really interesting. We start talking about dependency mapping, which I know this product can do, and we’re going to get into that a bit later. But one more piece that I’m curious about is Kubernetes. Is that like my Dev stood up a Kubernetes cluster and wants it to be managed and monitored? Is it like that or is it like mostly Tanzoo kind of stuff?

[00:09:16.080] – David
Okay. We have integrations with all Kubernetes, especially Tanzan Kubernetes grid with VMware, but it is a first class citizen. So if you want to look at your pods or containers down to the cluster level to kind of see all the information. So I mean, the relationship mapping, what cluster belongs to, what data center belongs to, all the storage and capacity constraints. So as you deploy, you have errors, or maybe you need to have some compliance and get alerting from that. You do get that relationship mapping. So if you need to troubleshoot something, you can actually see the root cause of that not just result, resolving or fixing that particular instance, for example.

[00:09:54.560] – David
So a lot of times you’re just looking at that moment in time, but you’re not really resolving what the root of the.

[00:10:05.350] – Ned
Network, right, Ethan? That’s always the answer.

[00:10:11.510] – Sujatha
Well, since he’s through about the network, let me also say that we do have integration with we realized network insight. So people are throwing things at eat and he can turn around and say, Let me show you. And so you can have the integration again, it’s inside of your apps where you can say, here’s the network flow that’s coming in, not my problem wherever the problem is.

[00:10:36.460] – David
Right.

[00:10:36.640] – Sujatha
So then it does help you to truly have that complete visibility, including network.

[00:10:43.800] – Ethan
And if you’re in the audience, we did a show on the Network Insight component of this. By the way, you can search on packetbushers. Net and find that.

[00:10:51.670] – David
So when we say unified across the entire data Center, I think we really mean unified across it. Right.

[00:10:59.940] – Ned
What I’d like to drill down to now is maybe some of the major use cases where people are rolling out or you think people might roll out vROps cloud, especially if they don’t already have VR Ops in their data center. David, what are some of the major use cases you’re seeing where this technology is being applied?

[00:11:18.590] – David
That’s one of the key things when we talk in the network, it’s really reducing that finger pointing when they’re pointing out. Well, like I said, that’s an application. I don’t have visibility into the VMs from the network team and vice versa the application. Well, everything’s looking good on my side. It must be some kind of network flow issues. But the key point of this is really to meet the modern engineer, the DevOps personas to your traditional It Ops, your sister admins and so forth. And also, what we’re seeing is you kind of have a mixture of your virtual admins and your cloud admins.

[00:11:52.370] – David
A lot of times the VMware guys or the infrastructure guys don’t speak cloud, and then vice versa, these modern cloud engineers. Well, I don’t really do this on Prem stuff. So things like integration with VMware cloud, for example, we have VMware cloud on AWS, we have Azure VMware services. We have GCBE Google Cloud VMware engine. So this kind of brings that VMware terminology, what you’re familiar with. So you can easily transition and deploy your objects. Now it’s VM containers and everything into the cloud. So you can understand that you don’t have to have that time to understand if you’re following the best practices.

[00:12:34.200] – David
So everything that you’re familiar with on Prem and vice versa, you’re going from the top down. If you’re familiar with a lot of the public cloud services, it’s easier to move some of the services down to on Prem, whether it might be data sovereignty issues or maybe security purposes or financial health care, that requires that these have to be on Prem. So kind of really speaks to both personas or multiple personas with this unified platform.

[00:13:01.190] – Ethan
Okay. So you mentioned reduced finger pointing as you kind of led into that bit. A lot of people that use monitoring software, they think about visibility software. They may go back in time and think about in terms of red light, green light. That would be a very shallow way to look at this. But can I use this product in that way, too?

[00:13:21.950] – David
I guess we’re one thing I want to kind of talk, AKS. Well, another name with integration we’re integrated with everything is we are integrated with the realist login site, right? So we say red light. Hey, you’re getting these alerts. Hey, do you live and die of this? Something dying. Or maybe it’s been like, hey, somebody’s logging on. Nothing is mad. Nothing’s bad here, but just a heads up. But basically, when we talk about this, when we talk about the integration with Login site cloud, it acts as a auditing tool, right?

[00:13:52.060] – David
Hey, anytime somebody opens and enters accesses VMs, or if there’s some kind of error that’s going to give you alerting, we have integrations with Slack with Pager duty. Right. So you can respond to everything in real time. So that’s one of the key things is that we have data retention requirements. So that’s one of the key things that US operations and all many integrations with all the complete public Cloud and VMware data center stack is to give you the ability to respond and see things in real time, whether it’s capacity, application issues or compliance.

[00:14:27.620] – David
We’re talking about compliance. You mentioned that’s the fourth value tenant of the self driving data center, and you can measure against your Vsphere. Hardening guidelines. You have six out of the box compliance templates, HIPAA, Pcima, Dice, ISO, and CIS. But again, if you have tons of other ones that you need to follow by, you can create these custom compliance templates as well. So when something drifts within that template, you get alerted, you’re sponsored and you’re back in the green light. No red light.

[00:15:00.110] – Ethan
All right, we get it. There’s a lot you can do here, and it sounds like this is one of those tools that once you get it stood up. And if you’re using it maximally, it becomes kind of a centerpiece of your operations. Now, the way you folks describe this product via vROps Cloud is around four pillars. And when we were talking about this prepping for the show, you described it to me as workload optimization, capacity planning, integrated compliance, which you were just mentioning, David, and also intelligent troubleshooting.

[00:15:29.480] – Ethan
Not just troubleshooting, but get some brains around it. So you can really drill into the problem quickly.

[00:15:34.240] – David
Intelligent. We call it intelligent troubleshooting. Everything goes wrong. So you don’t have to spend too much time to figure it out.

[00:15:43.390] – Ethan
So what we want to do going forward is get nerdy with it built around these four pillars. And that is let’s look at workload optimization, capacity plan and compliance and intelligent troubleshooting in more detail. The first one, then, is workload optimization. I want to understand what we mean by first of all, workload. What kind of workloads are we talking about? That I can deal with here VMs containers, others. How do you mean workload in this context.

[00:16:16.250] – David
When you say workload? I mean that’s everything. It could be your applications, your web service databases. It could be your native public cloud app, Kubernetes Times, for example. So anything that you can deploy in the structure, we are able to visualize it, we’re able to help manage it. And then ultimately, we can help optimize it. So we’re talking about performance optimization. This is driven by that predefined business and operational intent. So if you want to have everything completely balanced across your clusters and data centers, it will do that automatically.

[00:16:53.810] – David
Or if you want to do consolidation. Hey, fill up each data center, each cluster first before you start moving your work close to another one. So this is all done by that predictive DRS. That’s kind of the engine behind that.

[00:17:07.900] – Ned
There one of the more challenging things I think of in optimization is that what’s optimal for my organization and my application may not be optimal for your organization application. Maybe I have a requirement that I need 20% of overhead, and you like to run real leads, like, 99% to the hilt all the time. So what are we talking about when we say optimize? And how do we customize that optimization for the organization?

[00:17:39.210] – David
So that’s the thing you have these thresholds, like, exactly what you’re saying? Well, what is optimal mean for me? Well, I need to be optimal, because if I have one of my buddies in Philadelphia had a power outage, right. So the thing is being optimized. Okay, well, I need to make sure that I have at least 50% availability so I can move all my workloads over. And that’s one of the key things there. But another thing, a lot of other businesses, they don’t have that flexibility to run a 50%.

[00:18:10.200] – David
That’s expensive. I mean, you got to have another budget there. You have to run really lean. So that’s one of the key things maybe bring optimal is cost optimization. Right. So I have to make sure that, hey, I cross my fingers. Hope everything’s right. I need to maximize my current resources. And that’s one of the things we’re talking about consolidation. We want to make sure that we’re using as much as resources that we have before we start spending and deploying out more infrastructure. And that’s one another thing is where the SaaS solutions really kind of helps coming out.

[00:18:45.060] – David
So again, if we should run into this restraint. We can actually connect to the cloud. We can blow things out, and that’ll be our backup there without thinking about actually real estate, physical real estate that we have to expand and scale out.

[00:19:01.140] – Sujatha
There couple more points to that, because this is one of those really important things for our cloud admins and infrastructure admins as they deal with Netted as an important question, right? For each person. Optimization may mean different things, and you may be wanting to optimize for performance. Somebody may and the same person. Sometimes it’s all within the same cloud admin who needs to do all of this? They may want to optimize for performance for some of them optimize for capacity. And depending on that workload or that application, you may want to do your placement for what we call AKS VM happiness.

[00:19:41.990] – Sujatha
We want to make sure this critical VM is always happy. It’s always in the green state. So as we realize operations Admin, you come in and specify what those business intents are or those optimization intents are. Let me add this other thing that we have what is known as what if analysis. So you don’t even have to go do this before you try it out. So you can come into realize operations and run these what if analysis to see what will happen if I were to do this?

[00:20:21.660] – Sujatha
If I were to do my load balancing across this cluster, what will be my impact? So you can run such analysis so that you know what the impact is going to be. And then, as David was saying, you can automate it so that you can even say policies and go. If you see such a such threshold, go ahead and do this optimization for me. So the Cloud admin VI admin really has all the powers and the tools that they need to take care of these optimizations.

[00:20:56.310] – Ethan
So I want to know what kind of optimizations we can do. So things like maybe the VM needs more Ram, it needs more CPU or even GPU. I can do that kind of stuff, I assume. But then you also mentioned higher level functions like load balancing across a cluster where maybe I’d change how my load balancing is being done. Can you talk me through the sorts of optimizations that I might see?

[00:21:23.650] – David
Yeah, sure. So there’s two stories here, right. You have your VMs that are over provisioned. They get too much food, too much resources, and they’re the ones that other VMs that are resource starved. Hey, I need some nutrition. I’m not getting enough love and tender love and attention here. So that’s the thing that we talked about balanced all that stuff. Well, what US operations cloud can do is ask you and look at this. We’re wasting a lot of resources here on these VMs that are not using it.

[00:21:53.850] – David
They’re over provisioned. They’re giving up too much capacity. Or maybe they’re sitting idle. Right. So we can actually reclaim these Azure. We talking about orphan VMs. Everybody just deploying these VMs out. They use it for a while now. They’re orphans. So we can reclaim all those orphan VMs ones that have been powered on not being utilized and then give it these resources to the resource. So we talk about the noisy neighbor. For example, I’m sitting here, but I’m not doing anything. But you’re using all these resources so we can balance things out.

[00:22:24.880] – David
Another thing is when we talk about capacity optimization, too. Right. But we have this real time capacity analytics engine. It’s forward looking. So it runs through your data using machine learning, and it does pattern matching. It is pattern recognition and is able to quickly tell customers when they start running out of capacity. So it’s just this forecasting engine. So when we talk about time remaining, you have this graph. So based on historical data, how much utilization has been used. So that way, when you talk about end of year or future planning, how much capacity do we need to grow by?

[00:23:00.890] – David
So I’ll tell you exactly to the day, how much time before you run into capacity, and again, based on historical data collections, the more data that we have, the better, the smaller that forecasting band will be able to predict how much time we have available there.

[00:23:17.510] – Ethan
So focusing on workload optimization for a second. David Will vROps cloud. Will it recommend things that should be done? Or will it do things and say, like, hey, this was an over provisioned VM with a whole bunch of Ram. It never used. This other one needed it. I kind of moved the Ram. I changed the Ram allocations for your job done.

[00:23:37.750] – David
Yes.

[00:23:38.100] – David
So you get three options. You can schedule it, say, hey, every Friday night at 09:00 P.m.. Go ahead and do it for me. Or you can do it manually. I don’t want you to do it manually. Let me figure out whenever I want to do it. Or the third option is you can automate it. Hey, I see it. Let me do it for you. So you think about this teenager, right? How much do you trust your teenager to come home at a certain time? So that’s the thing.

[00:24:02.610] – David
Hey, go ahead. Come around this time. Let me give you a text. Hey, it’s an hour before you should come or full trust. They’ll be home when they’ll be home. It gives you that comfort of knowing that you can schedule it do on your own or do just do it for me. You see it. Do it.

[00:24:19.740] – Ethan
So what happens when the workload is sitting on a piece of metal that’s out of resources? It needs more. But there is no more CPU. There is no more Ram to give it. Can I do something like the motion that workload somewhere else or something like that?

[00:24:33.970] – David
Yeah, absolutely. And that’s what the performance optimization for. So you can set your predefined business and operations. So for number one, we’re talking about business intent. Okay, well, hey, this is all my database buses. What the heck is this web server doing here? You’re not supposed to be in this group. Get out. So I’m going to be motion these web servers that don’t belong in my database cluster, move it to the other web server clusters. So that way we can consolidate everything to meet Estlays, for example, and licensing requirements.

[00:25:12.750] – David
And now you have your operational sense of, hey, now that we have certain data centers and hosts, it’s supposed to be no more than 70% across utilization. So, hey, if it hits around 75, I see another one running out 40. Hey, let me balance you. Let me motion this automatically for you so we can meet these operational predefined intent.

[00:25:34.750] – Ethan
Now, can an intent be client sensitive? That is, maybe I’ve got a client that’s hitting a workload, and there’s a lot of latency. Does do you realize vROps cloud care about the client side, or is it focused on the metal and the infrastructure, so to speak.

[00:25:52.810] – David
The flexibility? You have a lot of flexibility to define your intention. So absolutely, it’s not based on the metal on the hardware. So again, we talked about you have Tags you have out of the box Tags, or you can create your own custom Tags, for example, say, hey, when we talk about resource utilization, operating systems, there’s a number of things that you can really automate this to really identify who and what belongs to what applications and how much resources that we have in this resource that is available for the objects.

[00:26:27.290] – David
And when I say objects, that’s all your applications and virtual machines and containers and everything in that inventory, you mentioned capacity.

[00:26:36.870] – Ethan
And that was another one of our pillars here. This the capacity planning component. Okay, what kind of capacity can vROps cloud keep tabs on? I think obvious things would be like CPU and Ram. But, hey, I’m a network guy. Can I keep track on network stuff as well?

[00:26:53.420] – David
So this has a lot to do with the network insight integration that’s going to tell you all the network flows and where you have the bottlenecks, where you have the resource contentions. And that way you’re going to get again, when we talk about this, you’ll get all these alerting. So if you need to reconfigure your virtual distributed switches or understand, have a deeper look into your network groups. This is where really, the network insight integration is going to help you understand that back end network flows and network intentions.

[00:27:22.450] – Ned
So, David, or suggest that I’m curious about how vROps cloud approaches cloud capacity, by which I mean managing capacity in the public cloud, because usually the problem is not a lack of capacity. It’s actually the inverse of using too much capacity. So what does that look like when you apply it to a public cloud provider?

[00:27:43.360] – Sujatha
It is pretty much the same AKS with. And as I mentioned, that is our intent so that our customers get the same experience regardless of what cloud that they are dealing with. So the capabilities that are available to our customers, be it reclamation of your workload or taking a look to do your right sizing of your workload. The experience is the same whether it is a vsphere environment VMware cloud environment or a public cloud environment. So kind of follow up of what Ethan was asking us earlier about the capacity planning and how do you do that?

[00:28:25.600] – Sujatha
One of the things that vROps does for our customers is to first and foremost, give you that visibility on what is happening with your cloud, whichever cloud it is, how much are you currently using? What is the base stage? What are the recommendations we can give you? So it starts kind of from there. Let me tell you what’s happening with your system and then for planning. There’s all kinds of what if analysis that I was talking about earlier, and so you could be taking a look to see if I was to expand and get more capacity.

[00:29:03.280] – Sujatha
What does that mean for me from a cost perspective, I know we didn’t touch on cost thus far, but that is. Another big portion of what we do is to let you do your planning both with capacity and the most important dollar numbers in mind. So you could do your what if analysis and say, let’s even take a migration use case. Right. If you have something private cloud and you’re wanting to move it to a public cloud and you want to do some analysis and say, what does it mean for me if I took it to Azure?

[00:29:36.200] – Sujatha
What does it mean if I took to AWS? What does it mean if I took to VMware cloud, we let our cloud admins do that analysis across the board. And so to kind of bring it back to the question you asked. Essentially, the capabilities that we have are for all types of cloud, including public cloud.

[00:29:59.050] – David
Another job I want to add to that. Is that the cool thing about this as well? Most important, how much is it going to cost me? So we do have scenario planning that you move your workloads, you’re looking at small, medium large templates that you want to deploy your application. So we actually have a calculator. It’s going to show you so pretty good cost. So as we move up and can we afford this and do we have budget for this? And most importantly, as we move the applications in this what if scenario to the public cloud, it all shows that, hey, how much resources that might freeing up.

[00:30:32.300] – David
So with that, can I go ahead and remove some of the old hardware when we talk about hardware refreshes or maybe how much more capacity we have on premise that we can actually deploy new applications out? So again, going from both sides and vice versa. If you want to move some of your workloads back down, do we have the resources available on Prem to support incoming applications? If you’re migrating from cloud down and vice versa. The cool thing about it is actually understanding team that calculator see how much it will cost if you’re thinking about migrating these workloads up into the cloud.

[00:31:06.740] – Ned
One of the things that’s really challenging about right Sizing is understanding context, right. Because once again, it comes back to the business context. What makes sense? And what does it mean to right size the application? Is there a way to express the business intent out to the recommendation engine? So it understands what right Sizing means for an application, or is it inferring right sizing through machine learning or something along those lines?

[00:31:33.810] – David
Okay, so the right sizing one thing, it’s kind of scary part, for example, you’re talking about their database admins personally, from my experience back then when I used to be It systems administrator, they want all the resources again just to make sure they have it for the spikes for the busy moments and the things. Again, through the analytics machine learning engine, it will factor in your average. It will look at your highs, it’ll look at the lows. So sometimes you might have these anomalies where you have these spikes, and sometimes that could be due to Rose and a part of the network.

[00:32:09.030] – David
Maybe something happened to the network there that caused a spike where something went down. But it really understands the historical data. And basically, when we do the right Sizing importantly, not just looks at the averages, but it looks at the highs. It looks at the time frames of how often that happens. So that way through the artificial intelligence engine in the back end will actually factor in how much resources it will take. But it’ll factor. And again, it does look at the historical data before it does the right size and recommendations, whether it’s scheduled or automated.

[00:32:43.670] – Sujatha
The heart of it is how the admins come and specify their operational intent and their business intent. So to the question you were asking earlier, too, some of them may be very application focused, so they can come in and specify these intents based on applications. Right. The way they define these groupings if you would, and these policies, if you would is truly, however, it is that they want to manage it, and we realize operations gives our users that flexibility. So they could take it from an approach of here are my applications, and I’m going to declare my operational intent and business intent based on my applications, or they could come in and say, well, at the end of the day, I’m just managing these infrastructures.

[00:33:33.910] – Sujatha
And so I’m going to specify these at the infrastructure level. So truly, it is how the flexibility is there for them to define these intents at the level at which they want to manage it so they can tell us, is it balance? Is it capacity? Is it cost? And then, as David was saying, we have our AIML engine kicking in and doing it for their intent based on all the observations we’ve seen in the infrastructure.

[00:34:04.250] – Ethan
Okay, so we’ve covered now workload optimization as a pillar, capacity planning as a pillar. I want to talk about compliance now, which is a third pillar here. In the top of the show we mentioned there were, I think you said six out of the box compliance templates covering a variety of regulations. Is that the only way to think about compliance here where we’re focused on regulatory compliance such as HIPAA and socks and so on.

[00:34:32.210] – David
So there’s a lot of things with our compliance there. So when we create, like, a custom compliance template for I’m going to use a real use case for Azure Storage, for example. So getting that connection with Azure Storage, one of the use cases where previous customers is that, hey, same thing like you get the alerts if you’re not using VM or VXLAN to get those integrated storage analytics things with the compliance we’ll talk about is if you have specific clusters that get a specific threshold. For example, right, when you’re becoming close to 80% of utilization, your compliance requires, whether for disaster recovery purposes or high availability, whatever it is that you have to abide by for your customers that hey, when you’re meeting or coming close within the specific thresholds, again, you’re going to get alerted.

[00:35:26.930] – David
Hey, you need to better manage your workloads. You need to consider adding extra node or think about reducing some of the orphan VMs or objects are not being utilized. It’s kind of a better looking to this. So then you Azure within the compliance. When we’re talking about capacity management, for example, to make sure that you don’t have any bottlenecks as they come up, because we’re really approaching that close to that red line that you should be aware of.

[00:35:56.640] – Ethan
Well, this goes back to the custom templating that I can do. Then I can set up a template that is arbitrary to whatever my bid. We’re using the word intent a lot today. So I’ll stick with that. Whatever my intent is for the infrastructure, I can set up a template, and then when it is no longer in compliance with that, the realized apps can flag me.

[00:36:18.410] – David
One of the things when you talk about the flagging, they all kind of ties back into something that with the login site within your operations cloud. So we’re going to get a flag, whereas you’re looking at it could be a number of things. Certain applications are making a lot of noise, and we’re talking about resource contention. They might be using a ton of CPU, or it could be a lot of Ram that’s being utilized. So looking at all these flags, for example, it will help you kind of stay within that kind of it will help you maintain that compliance to making sure that to meet your SLAs for your customers and use like this, or maybe have to better load balance application, or you need to spin up another instance to support these busy moments, for example.

[00:37:07.560] – David
So that’s one of the key things when we talk about the compliance part of that. There just from network utilization to application noise to CPU utilization across the VMs to either move in somewhere else. Or maybe it’s time to deploy another application, another instance to help manage these loads.

[00:37:31.550] – Ned
It sounds a lot like rules engine. Essentially, you’ve got some rules and some of those are created by vROps, and some of those you might be created by the administrator to express an intent or a rule you want to abide by and then some actions to take. Is the rule engine kind of the same across all these different pillars we’ve been talking about? Or is it a separate interface for compliance versus capacity versus the workload optimization?

[00:38:00.170] – David
You do everything unified in the real estate operations cloud, but you’ll have different dashboards. You can create a different dashboard. So that’s why you have multiple users and admins log in to your operations. So you can create these metrics, these dashboards and analytics that are specific to each team, so that when you log into it. Hey, I just need to look at the compliance part of your storage, for example. And then you have another page that’s going to be showing all the metrics and dashboards and analytics for my Kubernetes.

[00:38:31.490] – David
So that’s why you can create there’s a whole page that you have custom and out of the box metrics and dashboards. So that way you don’t have to create or define each page every time you log in. So it’s there. So you look at what you see and then same thing. If you need to correlate specific issues, you can look at all the dashboards. You can integrate it together. Kind of look at a specific time. Last week, around 01:00 p.m.. On Wednesday, we had an outage or resource intentions across the data center, so you can correlate these graphs, look at super metrics, and you can correlate that.

[00:39:07.140] – David
Hey, you know what? There are actually issues across the board from every one of these dashboards. So when you do this correlation so you can actually talk about the root cause analysis, you can look into that deeper and understand across board what might be happening. It could be a breach and environment. Again, it could have been some natural disaster outages that might have caused something or one of your generators went out and caused a cluster to go down, causing a lot of more heavier load and everything else to manage during that time.

[00:39:38.990] – Ned
Okay, so I’ve got a rules engine that’s churning through all this, but the way that’s presented to different personas will depend on what that person is interested in. If I’m a compliance person, that’s the thing I want to do. That’s the view that I’ll get assuming that something does happen with compliance, is there remediation that can automatically be applied to a compliance condition, or is it something I’m just getting notified and then I AKS, the compliance person have to go in and make that change or fix that issue.

[00:40:09.830] – David
The awesome thing is AKS I mentioned earlier. This is applicable across the board. So you have three settings. Just alert me and I’ll do it on my dedicated maintenance window that I can only update and make changes to or I can schedule it. Hey, again, this is our automatic scheduler. This is the time that we’re just doing. I trust it. Just go and do it at the specific time, and then the third one is the automated one. Say, hey, if you see it, do something, fix it.

[00:40:39.910] – David
I want it fully automated. I don’t want to do things. Just see it. Give me the alert. I can look at it. And I already know I have comfort knowing that it’s already been remediated automatically through the AI Ops engine. There.

[00:40:50.740] – Ethan
Now, is there an auditor friendly report that I could generate for certain regulations that I might be getting audited for?

[00:40:56.880] – David
Yeah. So like I said, everything that happened, the events and logs and actions, they are all tracked. So end of the week, if you need to show networking or security auditing, it will show you exactly what’s happened, what actions were taken toward timestamps. So again, it’s full tracking to do weekends or monthly auditing, see what happened. And again, just to really improve on the best practices, understand the weakest link in the chain there to understand what you need more visibility and attention to.

[00:41:31.780] – Ethan
Okay, all right. We are ready for the fourth pillar, and this is my personal favorite. So I’ve saved this one for last. Intelligent Troubleshooting because I have always been a really big fan of using some kind of a visibility tool to help me get to the bottom of what’s going on without me having to go on instinct and start logging into things and checking stuff. I want the tool to tell me so I can take all that time off the table and get closer to the problem and get it fixed more quickly.

[00:41:57.880] – Ethan
So. Okay, if VR Ops cloud is going to be that Intelligent Troubleshooting partner tool if you will, so that I don’t have to log into all the things. Let’s start with this. What is the tools view of my infrastructure? And here’s how I mean that we’ve been talking contextually hybrid cloud in this discussion. We’ve been talking about Kubernetes. I’ve got workloads and stuff everywhere. I might not even know because of the complexity of this infrastructure, where the apps are, where a Dev pushed code to today, what the dependencies even look like until something breaks?

[00:42:38.910] – Ethan
What I’m hoping you’re going to tell me here is how VR Ops cloud fixes that scenario. For me. It makes it easier for me to solve my problems when they come to me because things are broken.

[00:42:49.260] – Sujatha
In the beginning of this talk, I started off by saying this is a self driving operations, right? So this is the heart of what vROps does. And we are talking about a really powerful AI and Ml engine that lets us do this really powerfully. I wouldn’t even want you to consider this as a partner tool, but this is the tool for folks to do their troubleshooting and their remediation. And as David talked about earlier, it’s really again intent driven it’s up to you whether you want to do it fully automated or you just want alerts and you take care of it.

[00:43:33.050] – Sujatha
But we are is really good at putting that AI Ml engine to use and do things like even predictive forecasting for you. Right. So we use I know you said this is listened to all of our tech people. So we use a discrete and fast furious transformation algorithm for real time forecasting, and we apply it across the board. I remember Ned asking us a question about the things that I do is it across the various capabilities, and it is we use these predictive forecasting for workload placement for capacity analysis for right Sizing and predictive DRS use cases.

[00:44:18.500] – Sujatha
So this is like the heart of the V Realize Operations product. And we also do root cost analysis. You start off by saying I might have many, many things that are impacting. The output will Vir up, help me understand what’s going on. And so this is the heart of it. Right? We do the root cost analysis based on dynamic thresholds based on anomalies. There are things that we, as humans may not be able to pick out, but the machine can. So even if there are some anomalies in what we are noticing across the boat, we can highlight it for you and flag it for you again, up to you on whether you want this to be an alert.

[00:45:03.650] – Sujatha
Do you want it to be auto remediated? But this is something that we do for our users. And I’ll touch on the third one because we talked about networking earlier. That’s the other piece of it.

[00:45:15.990] – Sujatha
Right.

[00:45:16.230] – Sujatha
The Aimlbased analytics for network outliners are also in place so that you pull in all of that information and it’s not in Silos. We get this holistic view across all of these various factors and help our users to get to the bottom really quickly.

[00:45:35.880] – Ned
So walk me through a potential issue and how VR apps would be used to address that. So let’s say I’ve got an application. It’s running on BMC it’s up. I can maybe get to the front page, but it’s running real slow. Users are complaining. I’m getting help desk tickets coming in. How do I leverage vROps cloud to help me get to the root of what’s going on with that application.

[00:45:59.990] – Sujatha
Sure. A couple of things. Let me introduce you to something that we are bringing in the very near future. Synthetic monitoring. This is a big deal for our users to start doing synthetic monitoring, for example, for the use case specifically, Ned, that you said it’ll just bring that much more powerful resolutions for our users. It’s coming. It’s not there yet. Just again, a plug. Today we realize Operations Cloud SRE team uses this technology and so we are bundling it up and bringing it to our users to do the synthetic monitoring.

[00:46:39.180] – Sujatha
So something your specific use case. Synthetic monitoring will be helpful even before the problem shows up. Right. You can go and set in and say, here’s the way my application is used. Here’s the workflow. So I want to go test it out. So that’s coming. But even currently, today in the Troubleshooting workbench, you can go and set up the workflows that you number one, post a problem occurring. You can go and say here’s how I go about looking at this problem. And I want you to record this for me and then make this part of what you wear.

[00:47:18.570] – Sujatha
Vrops keeps track of. Right. Then I don’t know if you have the time to do justice to the Troubleshooting workbench, but that is a powerful tool. We introduced a couple of releases ago that will help our admins to double click into such problems and get ahead of it. And I’ll speak on one other thing. And then I’ll see if David wants to add again on the cloud, we introduce real near time monitoring. What else means is currently on Prem product. For all of our Vsphere based clouds.

[00:47:55.160] – Sujatha
We collect metrics every five minutes. Now, that may be sufficient for 80% of the use cases. There are some where you really want that collection to be much quicker. Right. And so we introduced something on the cloud side where you can go and say for this workload, I really want five minute intervals so that I can double click and you can even adjust it to 22nd for short duration. Right. Because otherwise you will be buried with metrics and data. So there is ways to adjust that knob depending on how critical the issue is where you can see near real time monitoring data and make those adjustments.

[00:48:36.730] – Sujatha
So let me see, David, if you wanted to add anything more to that.

[00:48:40.750] – David
Yeah. So with this awesome thing is when we talk about the complete integration because we’re using and needing information from everywhere to really get down to Troubleshooting, we have the full integrations with Skyline and that’s that Proactive intelligence, and we have the root cause analysis with viewers login site too. And then that’s what we kind of form and visualize everything through viewers. Operations Cloud. So let me talk about it. So maybe first you’re going to get this explosion of logs in your environment. So what it is is when we have this automated root cause analysis approach, it’s able to detect and surface relevant logs in terms of log clusters.

[00:49:21.060] – David
And then it’s going to present us in a way where you can see this in real estate operations. So that way you’re only looking at the reduced number of logs. It’s going to show you only the significant ones that might be the potential root cause. Another thing is, of course, we can have integration with the Skyline as well. So again, productivity and reliability. Vmware it receives data from the Skyline Collector, so it performs all its analytics. You align it with the VMware best Practices and KB article.

[00:49:52.730] – David
Aks Security Advisories, and it’s able to really help identify these problematic areas that can be addressed with a recommended solution or an automated solution. So you think about everything from starting from a simple checking of data, looking at your configuration attached levels and also all the way to the complex cross cloud checks as well. So when we talk about this, everything from originating from a log to visualizing through the Intelligent Dashboards and super metrics and all the way up to the KB Insights, where you’ll have the Proactive intelligence with Skyline to really tie everything together, to really help resolve and reduce that.

[00:50:31.600] – David
Meantime, the resolution there and again, looking at everything not just specific problems but the object relationships that are made from the pods, the clusters to the virtual machines, the cluster to the data center, to the underlying hardware it’s sitting on. Or again, or it might be up into the public cloud that you might have deployed these modern cloud apps.

[00:50:52.080] – Ethan
And folks, I love management tools. I really do. And I appreciate the discussion here about what all vROps cloud does. Got to get that product name right. Vrops Cloud good stuff. Anything that helps me visualize what’s actually going on in my infrastructure and get to the bottom of problems do capacity to management. Fantastic. Now I know there were some announcements at VMworld about enhancements and so on. So, Jonathan, can you give us a caption on any new and exciting things that we didn’t cover today?

[00:51:26.990] – Sujatha
Sure. Let me give you a couple of them. There’s too many, but I’ll give two. One is that we have really redesigned that experience with viralized operations. So if there are users who have been using VR for many years now, I want you to know that we are focused on simplifying your experience. The product is so Azure, but we want to make sure that you’re maximizing on what you’ve already invested. So what you’ll see the number one thing as soon as you log in once you’ve upgraded or if you’re on the cloud we’ve upgraded for you is that experience has become new and so you can maximize what you already have in hand.

[00:52:10.380] – Sujatha
And the other one I lead in is the point that I was making earlier about us really wanting to give you a differentiated, integrated experience as you consume more and more of VMware service mesh. If today you are already using VMware Cloud on AWS, we have introduced a one click activation for V realized operations. Truly, with just one click from your VMC on AWS, you can start experiencing cloud management and you get a 30 day free trial. David can speak more to it, but it’s at your fingertips.

[00:52:51.020] – Sujatha
You already have BMC on AWS. One click. You can start videoized operations and start managing your environments.

[00:52:59.000] – David
Check out our landing page. We do have the 30 day trial. Within minutes, you can spin up your own instance, start margin your objects and see how everything works. It doesn’t take any time to deploy anything to configure everything again. Just get that day to experience out of the box. And if you’re curious to understand how to click around, we do have our Pathfinder VMware dot com page that we have tons of level One, two and three where you’re going to have articles and blogs, you have videos, and we’re going to have hands on labs that they’ll walk you through.

[00:53:32.420] – David
Really that Day two experience. So when you click that trial, you know exactly what to do. Look at the dashboards and see what’s out of the box and how to configure your own custom dashboards and metrics for visibility.

[00:53:44.000] – Ethan
Fantastic. Thanks to both of you for appearing on Day Two Cloud. Thanks to VMware for sponsoring. That is how Ned and I feed our family. So we do appreciate our sponsors. If you are out there listening and you end up going to VMware and giving V Realize Operations Cloud a try. Hey, tell them you heard about it on Day Two Cloud and the Packet Pushers Podcast Network. We’d appreciate that virtual high five to you for tuning in. You are a wonderful, wonderful human. If you have suggestions for future shows, products you want us to cover vendors you want us to talk to.

[00:54:11.110] – Ethan
Maybe you’re an expert in something and you want to talk about your cloudy experience. We would love to hear all about that from you. You can hit Ned or I up at Day Two Cloud show on Twitter or go up to Ned’s website. nedinthecloud. Com. You can fill out his contact form there. Did you know that Packet Pushers that the Packet Pushers Podcast Network has a weekly newsletter. Human Infrastructure Magazine. Human Infrastructure Magazine is loaded with the very best stuff that we have found on the Internet, mostly technical engineering stuff.

[00:54:36.570] – Ethan
So you can be smarter about your job from a week to week. Plus our own feature articles and commentary. It is free. It doesn’t suck. We don’t use your contact info, we don’t sell it to anybody or anything. Just go to packe pushers. Net newsletter. Sign up and you’ll get the very next issue. And until then, just remember, Cloud is what happens while it is making other plans.

More from this show

Day Two Cloud 147: Google Cloud Is Not Just For Devs

Today on Day Two Cloud we peel back the curtains on Google Cloud with a GCP insider to find out how Google Cloud differentiates itself, its embrace of a multi-cloud approach, and more. Our guest is Richard Seroter, Director of Outbound Product Management...

Episode 123