View Recording: Kicking VMware to the Curb – Unlock the Full Potential of Microsoft Solutions

So we are starting our conversation today about how to kick VMware to the curve. 0:0:18.730 –> 0:0:23.550 Concurrency That’s maybe a little bit of a spicy, spicy title for the session. 0:0:23.560 –> 0:0:32.70 Concurrency What really, really talking about is how you consider the idea of leaving or moving or exiting your on premise data center. 0:0:32.80 –> 0:0:36.70 Concurrency Or maybe augmenting your on premise data center or you’re trying to scale it back. 0:0:36.310 –> 0:0:39.830 Concurrency We’re going to talk a little bit about how companies are accomplishing them. 0:0:40.300 –> 0:0:48.480 Concurrency This how thinking about the picture and then what you need to think about as you start to examine your cloud data center ecosystem. 0:0:48.620 –> 0:0:51.340 Concurrency So with me today, I’m first I’ll just introduce myself. 0:0:51.350 –> 0:0:52.150 Concurrency I’m Nathan wisnoski. 0:0:52.160 –> 0:0:54.0 Concurrency I’m concurrency chief technology officer. 0:0:54.630 –> 0:0:59.760 Concurrency I lead many organizations in helping design their cloud journey. 0:0:59.770 –> 0:1:13.270 Concurrency How they look at the ecosystem they’re running today and how they Orient themselves toward a future that is allowing them to build on the cloud ecosystem of that’s available to them and I’ll be helping lead this conversation. 0:1:13.280 –> 0:1:15.290 Concurrency But with me is Aman Aman. 0:1:15.300 –> 0:1:15.510 Concurrency What? 0:1:15.520 –> 0:1:16.780 Concurrency You just introduce yourself a little bit. 0:1:18.130 –> 0:1:34.560 Ayman Mageed Hey everyone, Amon Mageed I am coming into concurrency as the new cloud practice leader bring in about 15-16 years of cloud and networking background and really looking forward to coming in with Nathan here and demystifying Azure to the masses. 0:1:36.230 –> 0:1:36.790 Concurrency Awesome. 0:1:36.800 –> 0:1:37.460 Concurrency Thanks amon. 0:1:38.610 –> 0:1:39.30 Concurrency OK. 0:1:39.130 –> 0:1:51.200 Concurrency So we’re gonna start this with just talking about a little bit of a self identification of where each of us might be at and associated with this self identification. 0:1:51.210 –> 0:2:7.440 Concurrency I think one of the better tools to use to think about that is a cloud maturity curve and this is one that I’ve been using for probably, oh, man, almost 678 years now in terms of how to think about where you are on a journey and you still may be anywhere on this cycle. 0:2:7.820 –> 0:2:10.610 Concurrency But I think it’s a great way for us to Orient where you are today. 0:2:11.240 –> 0:2:26.650 Concurrency So lower left hand corner which you can see is identification with organization that is still primarily on premise and oddly enough I still run into a lot of Organism which is completely there. 0:2:26.720 –> 0:2:32.320 Concurrency They’re using on premise, collaboration communication tools mostly on premise data infrastructure. 0:2:32.330 –> 0:2:37.730 Concurrency Maybe there are core application is running mostly on premise and it’s leveraging sort of historical technologies. 0:2:37.740 –> 0:2:39.550 Concurrency Perhaps it’s even iseries? 0:2:40.20 –> 0:2:51.130 Concurrency Kind of like bone anchor cake technologies, but also might be something which is more akin to like Windows virtual machines that are running the predominant business applications that exist in the ecosystem. 0:2:51.280 –> 0:2:55.910 Concurrency So many organizations still sort of sit at this legacy on premise, please. 0:2:56.280 –> 0:3:19.240 Concurrency And usually when an organization is there, there’s other parts of the organization that are starting to go around it and it’s have been going around it and perhaps for a long time, lighting up SAS applications, starting to build their applications in the cloud and exists in this sort of stagnant it, but scattered cloud that’s gotten bursted up within the business to be able to enable that new innovation. 0:3:20.250 –> 0:3:32.480 Concurrency What I’m finding is that as organ donations look to jump from one of these two spots and I’m one spot or the number two spot, they are thinking about what that future looks like for them. 0:3:32.610 –> 0:3:38.140 Concurrency And understanding that there’s a skills gap, there’s an opportunity for them to take advantage of cloud computing. 0:3:38.150 –> 0:3:40.420 Concurrency But the people may not be quite ready for it. 0:3:40.490 –> 0:3:45.980 Concurrency The skills may not be quite ready for it, and they’re not sure how to work with the economics as part of that picture. 0:3:46.170 –> 0:3:57.590 Concurrency So what we’re going to be talking about today is what do you need to think about and so why and what do you think about as you start to picture the movement up into these higher levels of maturity? 0:3:57.840 –> 0:4:6.200 Concurrency And one of the core elements that we’ll discuss today is what does it mean in the operationalized different cloud estate positions. 0:4:6.330 –> 0:4:14.560 Concurrency So as you look toward, maybe I could leave that colon and go to Azure, maybe I can not renew my on premise data center racks and move those into Azure. 0:4:14.570 –> 0:4:18.140 Concurrency Maybe I cannot renew some of that VMware contract and move that into Azure. 0:4:18.220 –> 0:4:23.810 Concurrency What does that appropriate operationalized cloud environment look like and how might I consider that picture? 0:4:26.320 –> 0:4:48.630 Concurrency And then in conjunction with that, how we think about what builds on top of it from building streamlined cloud product development and starting to leverage capabilities that are already being asked for by the business, artificial intelligence data estate, building, a modern application, all of these build on the skills that you’re establishing as part of an operationalized cloud as well as other elements that you might modernize along the way. 0:4:48.640 –> 0:4:54.380 Concurrency So we’ll talk about some of those parts of the picture and help you to think about how you might consider this in your own organization. 0:4:55.630 –> 0:5:3.890 Concurrency One thing I’m going to call out here as we start to talk about this is we’re gonna assertively use the chat as part of a Q&A session. 0:5:3.990 –> 0:5:4.670 Concurrency So as we go. 0:5:6.600 –> 0:5:10.50 Concurrency Long here I I want you to make sure you’re putting your questions in the chat. 0:5:10.160 –> 0:5:16.570 Concurrency We will be answering those as we go along, so I’ll either answer them out loud or email answer out loud or we’ll answer them in the chat as we go. 0:5:16.640 –> 0:5:20.660 Concurrency So drop those questions or comments that you have out there and we’d be happy to address them as we go. 0:5:22.120 –> 0:5:32.430 Concurrency OK, so as you think about this idea of like, could I leave this on from his data center or could I enable my new workloads to go into a cloud estate? 0:5:32.780 –> 0:5:39.430 Concurrency I want you to think about three different operational states that you might be living in at any given point. 0:5:39.980 –> 0:5:42.990 Concurrency The first is horizon one and horizon one. 0:5:43.0 –> 0:5:47.760 Concurrency Just kind of think about is this is was right in front of you. 0:5:47.770 –> 0:5:48.960 Concurrency This is what I’m doing right now. 0:5:48.970 –> 0:5:54.740 Concurrency This is how I have to operate the environment that I currently have that it’s just keeping the lights on. 0:5:54.750 –> 0:6:18.130 Concurrency Kind of activities and operating on premise estate, you have monitoring tools, backup tools, skills and in most on from substates you’re operating the sand and the networking and the storage and the VMware and the deployment in that environment and the ITSM tools that circle around it and the monitoring and backup and recovery, all of that is part of operating that legacy data center operations estate. 0:6:19.130 –> 0:6:27.720 Concurrency And when you start to look at the opportunity to move these into the cloud, realize you’re moving to a different horizon of operational excellence. 0:6:27.970 –> 0:6:41.220 Concurrency So when you move to Horizon 2, Horizon 2 is really defined as I have a virtual machine just like I had in my on premise environment, but that virtual machine is operating in a modern operational state. 0:6:41.390 –> 0:6:49.250 Concurrency So that means that I’m using cloud centric monitoring tools, cloud centric recovery tools, cloud centric Dr Tools, my deployment. 0:6:49.500 –> 0:6:51.890 Concurrency The system might be using infrastructures code. 0:6:52.20 –> 0:7:10.650 Concurrency I might be micro segmenting the workloads using a different kind of model that I used on premise, so it represents a skill jump as well as an operational change from your on premise environment, but also represents a way for your team to modernize their skill sets to be able to work in the Horizon 3 ecosystem. 0:7:10.860 –> 0:7:26.40 Concurrency So the Horizon 3 ecosystem, meaning cloud, native deployment, cloud, SaaS oriented deployment of data states or something like an avd or W 365, we’ll talk about later the Horizon 3 state represents this refactoring of the workload. 0:7:26.150 –> 0:7:29.670 Concurrency In some cases you can jump right from one to three. 0:7:29.810 –> 0:7:43.840 Concurrency In other cases, it makes a lot of sense just to move to Horizon 2 in order to move to horizon three, so we’ll talk about each of these, but realize that each represents an operational state that is different from the one that you’re operating in horizontal 1. 0:7:45.450 –> 0:7:57.0 Ayman Mageed I think this is one of those situations where us as an organization in many organizations have moved to Horizon 2 and we were very excited because back in the day we made the move into cloud. 0:7:57.190 –> 0:8:2.420 Ayman Mageed But then if you stopped there, there’s almost this law of diminishing returns, right where? 0:8:2.490 –> 0:8:10.120 Ayman Mageed Yes, we can take advantage of some of the cloud aspects, but we’re not really getting into Azure functions, serverless opportunities. 0:8:10.130 –> 0:8:21.960 Ayman Mageed We’re not managing containers and so that cost savings that we were looking for doesn’t materialize because we’ve taken some of our problems right into Azure with us and we haven’t solved for them, right. 0:8:21.970 –> 0:8:37.660 Ayman Mageed So if we this relates to this Broadcom move right and VMware like if we’re bringing this in, do we have a plan or are we just taking kind of the old into the new and just maintaining it at a potentially higher price as well? 0:8:38.410 –> 0:8:39.420 Concurrency Yeah, great point. 0:8:39.430 –> 0:8:39.870 Concurrency Great point. 0:8:42.250 –> 0:8:58.400 Concurrency So I think a good place to start in terms of how people are thinking about those migrations is not just the why, but also what is causing them to to consider moving out of their on premise ecosystem. 0:8:58.410 –> 0:8:59.110 Concurrency What are the like? 0:8:59.120 –> 0:9:0.410 Concurrency What spurs that on? 0:9:0.480 –> 0:9:3.130 Concurrency What causes that initial that initial perspective? 0:9:3.780 –> 0:9:7.350 Concurrency I think the one we see the most is when companies are going through and have a. 0:9:8.0 –> 0:9:14.30 Concurrency So when companies are going through mergers and acquisitions activity, there’s two situations that could be happening. 0:9:14.40 –> 0:9:24.140 Concurrency One is a company might be intaking, a company that they just bought into their new organization, or they might be spinning off a business into its own independent business. 0:9:24.150 –> 0:9:25.160 Concurrency They might be taking something. 0:9:25.170 –> 0:9:52.270 Concurrency They already are, and splitting it into in those projects, companies rarely decide to keep their current data center operations, and there’s two reasons why the first reason is because if it’s an intake that existing data center represents a high degree of risk, you probably already know organizations that have been compromised as a result of ransomware or security state problems in organizations that are in taking another organization into their state. 0:9:52.350 –> 0:9:57.200 Concurrency Rarely want to take on that risk, and they very rarely just want to lift them into an on premise data center. 0:9:57.250 –> 0:10:16.600 Concurrency Typically, we’re finding that we’re taking that established environment and we’re migrating that into a modernized horizon to uh data center in the cloud, A to mitigate the security risks because there’s a variety of things we can do as we make that migration and B, because it allows them to sort of clean up the operationally ecosystem along with it. 0:10:16.990 –> 0:10:18.800 Concurrency So that’s sort of one reason. 0:10:18.810 –> 0:10:23.640 Concurrency The second reason is because it establishes an operational state. 0:10:23.650 –> 0:10:32.460 Concurrency That is, that’s moving their staff toward a modern set of skills, and that’s also one of the reasons why PE firms tend to like this, especially when you’re spinning them out. 0:10:32.970 –> 0:10:38.340 Concurrency Is they also don’t want to be left with like the tech debt associated with that on premise environment. 0:10:38.410 –> 0:10:41.900 Concurrency They wanna bring that into a modern environment that they can hire with modern skills. 0:10:42.180 –> 0:10:46.260 Concurrency So very rarely do we find that companies wanna stick around with that on premise ecosystem. 0:10:46.270 –> 0:10:52.740 Concurrency They tend to want to mitigate the standing risk by moving it in into an environment that they can better support. 0:10:53.290 –> 0:10:59.740 Concurrency The second our sort of related ecosystem that kind of Spurs us on is I’ve got to Tier 2 data center. 0:11:0.10 –> 0:11:9.260 Concurrency My contract is near renewal or approaching renewal and I wanna look at a higher maturity vehicle to be able to operate my legacy workloads. 0:11:9.450 –> 0:11:15.530 Concurrency And I’ve heard, you know, some CIO say, like, there’s no reason to move legacy workloads to the cloud. 0:11:16.560 –> 0:11:19.150 Concurrency I I just 100% disagree with that. 0:11:19.700 –> 0:11:22.310 Concurrency Main reason being skill development. 0:11:22.320 –> 0:11:33.690 Concurrency Second reason being, establishing an ecosystem that those workloads can live in, that cleans up a lot of the Tier 1 Tier 2 mistakes that companies have had as they’ve been operating those workloads. 0:11:34.40 –> 0:11:45.440 Concurrency And the third reason for my perspective is looking for cost savings like I can generally find increased cost savings over a Tier 2 data center by moving it to Azure and then establishing reserved instances along with it. 0:11:46.280 –> 0:11:53.50 Concurrency So something to think about, uh, there, the third scenario is just simply on premise cost replacement. 0:11:53.140 –> 0:12:4.50 Concurrency If you have an existing data center and you’re starting to need to replace pieces of that, I need to buy a new data domain or I need to add storage to my san or I need to all those elements. 0:12:4.480 –> 0:12:15.900 Concurrency That is bad money, like you’re adding bad money rather than good money, because I can’t reduce that later I bought another hardware boat anchor that I’ve got to hang on to for X period of time and depreciate. 0:12:16.250 –> 0:12:17.920 Concurrency How do I get out of that cycle? 0:12:18.190 –> 0:12:20.820 Concurrency That is where I’m seeing many companies make that choice. 0:12:21.330 –> 0:12:28.540 Concurrency On the 4th scenario, that’s causing people to make that choice is they’re actively moving to SaaS and Paas models anyway. 0:12:28.550 –> 0:12:43.40 Concurrency So like great example, we’ll talk about a second is like if you have a big RDS or Citrix Farm, you’re probably already looking at moving to Azure Virtual desktop or W 365 it just even if I was moving it to the cloud, I wouldn’t lift those RDS servers. 0:12:43.50 –> 0:12:54.140 Concurrency I would be moving to one of those two models, and you’re probably already in the process of doing that, or you might be building new applications that are ground up container based apps which are just simply not going to be doing in your on premise environment. 0:12:54.150 –> 0:13:21.630 Concurrency For the most part, and that kind of goes to this point, which is the high percentage of new things are already landing in the cloud, the number of companies for example, that are putting their data state in on premise because system is like nothing it’s it’s mostly either a cloud snowflake Azure state or it’s going toward fabric as a SAS based estate that’s being built in the cloud, none of which renews the on premise data center as an option team in any perspective on these. 0:13:23.740 –> 0:13:44.820 Ayman Mageed No, I think we’re gonna hit on a topic here in the next slide, but really what it comes down to is, umm, are we making decisions because someone just notified us and told us where we need to be out of this by X date or are we actually taking into account the future growth and strategy of whether it be a cloud strategy or the enterprise strategy. 0:13:44.830 –> 0:13:56.480 Ayman Mageed So when we think of cloud as a commodity, we run into these situations where we do a quick business case, we move into cloud and then we are asking the question a couple months down the road. 0:13:56.490 –> 0:14:11.210 Ayman Mageed While we’re not saving money, but when we start looking at these migration options that we’re about to to pivot to, it really makes sense like I would not clean my garage out and keep the things that I should probably throw out or or modernize, right? 0:14:10.580 –> 0:14:10.820 Concurrency Yeah. 0:14:11.220 –> 0:14:16.890 Ayman Mageed And then put them back in, because then I I’ve just continued to waste that space and now I’m paying for it. Figuratively. 0:14:16.900 –> 0:14:25.750 Ayman Mageed Luckily, I’m not paying for my garage anymore, but that is that is the premise here is like when we are moving into cloud. 0:14:25.760 –> 0:14:39.730 Ayman Mageed We’re also shifting into that new mindset of how do I get rid of legacy datacenter operations that we’ve all held onto because we majority of us have grown up in, in that space and and and leveraging some of those kind of accelerators. 0:14:40.870 –> 0:14:44.40 Concurrency I love that example of a garage. 0:14:45.580 –> 0:14:50.150 Concurrency I recently had the most freeing experience I brought. 0:14:50.160 –> 0:15:12.250 Concurrency I got 2 dumpsters and I parked them outside of my house and I filled them up with junk that I’ve been caring from house to house with me and it was the most free experience that just Chuck stuff and I was having this conversation with my wife and she was like, oh, you’re gonna need like a week to like figure out what you’re gonna throw out in and organize all that. 0:15:12.260 –> 0:15:13.620 Concurrency I’m like I need a day. 0:15:14.930 –> 0:15:15.280 Ayman Mageed Yeah. 0:15:15.0 –> 0:15:16.160 Concurrency I need a day. 0:15:16.220 –> 0:15:30.90 Concurrency I know I I have no problem feeling 2 dumpsters with junk from my basement and just, you know, things in your garage and old wood that you have been hanging on to and all bikes that you wanted to repair. 0:15:30.100 –> 0:15:40.420 Concurrency Just junk and this is one of those free moments of my misspelled decommission to really hit on this idea. 0:15:40.630 –> 0:15:47.530 Concurrency Now one of the best things you can do at this point is to identify what you’re not moving, that there’s workloads. 0:15:47.540 –> 0:16:3.520 Concurrency That you can straight up, shut down and that’s a very freeing experience of exiting the data center is having an honest conversation about what you don’t need to move when we’re doing many of these versions and acquisitions projects. 0:16:3.970 –> 0:16:7.640 Concurrency One of the things that comes into play is a list of applications. 0:16:7.650 –> 0:16:13.460 Concurrency And let’s say there’s 300 applications with 500 servers tied to it, and that’s the like starting list. 0:16:13.750 –> 0:16:22.150 Concurrency And what happens very quickly as you get into that conversation is you discover, well, a third of them are utility servers that we need to operate. 0:16:22.160 –> 0:16:44.80 Concurrency The on premise data center and then another 20% are RDS and VDI systems and another good 10% of that is just stuff we can just shut off because nobody uses anymore or it’s not valuable the organization and you chop down that estate to this much more small subset just by decommissioning. 0:16:38.460 –> 0:16:38.830 Ayman Mageed Correct. 0:16:44.940 –> 0:16:46.180 Concurrency I’m in the yeah. 0:16:45.150 –> 0:16:54.350 Ayman Mageed I’m looking at a list of applications right now, and some of those naming conventions, even of the servers themselves, say deprecated decommissioned, but it’s still there, right? 0:16:54.360 –> 0:16:58.510 Ayman Mageed So it’s it’s literally another chance to do that cleanup of, yeah, I don’t. 0:16:58.520 –> 0:17:0.910 Ayman Mageed I’m pretty sure that server is not supposed to be there. 0:17:0.980 –> 0:17:6.10 Ayman Mageed We can, you know, have that conversation, not even put the effort into into migrating. 0:17:7.620 –> 0:17:17.510 Concurrency Curb set and then the flip side is this picture of migraine assess like you may be already really close to not even migrating that workload to begin with. 0:17:18.280 –> 0:17:20.370 Concurrency It’s tied in with the decommission, right? 0:17:20.380 –> 0:17:21.850 Concurrency You lit up a SAS platform. 0:17:22.420 –> 0:17:25.50 Concurrency Maybe you’ve been an example of this like I was talking to a company. 0:17:25.60 –> 0:17:31.450 Concurrency They had this like SharePoint 2000, 2010 or something SharePoint environment. 0:17:31.460 –> 0:17:35.820 Concurrency That’s like an on premise SharePoint environment and they’re on Office 365 is like, why do you still have that? 0:17:33.860 –> 0:17:34.90 Ayman Mageed You know. 0:17:36.680 –> 0:17:38.210 Concurrency It’s not patched, not update. 0:17:38.220 –> 0:17:40.570 Concurrency It’s like, oh, we just had some old stuff out there. 0:17:40.580 –> 0:17:44.110 Concurrency We’re trying to keep around like archive that thing, man. 0:17:44.200 –> 0:17:45.100 Concurrency Shut it off. 0:17:45.80 –> 0:17:46.680 Ayman Mageed At Penney’s. Yup. 0:17:45.110 –> 0:17:48.710 Concurrency Throw it in archive storage, get it? 0:17:50.720 –> 0:17:53.670 Concurrency Umm, so then this leaves us with like the three middle options. 0:17:53.680 –> 0:18:0.420 Concurrency OK, so leaves us with these and this is maybe where I’ll diverge a little bit from where some messaging you might get. 0:18:2.320 –> 0:18:9.610 Concurrency I would urge you to migrate if you’re moving to the cloud to think about these two options. 0:18:11.660 –> 0:18:15.570 Concurrency Migrating to native Azure virtual machines and modernizing workloads. 0:18:15.580 –> 0:18:23.40 Concurrency Along the way, I find that when companies move to AVS, they are not really achieving horizon too. 0:18:23.250 –> 0:18:27.160 Concurrency They’re like they’re they’re they’re sort of tempted by that idea. 0:18:30.190 –> 0:18:49.20 Concurrency I can maintain my current skill set and exit my days, but the problem with that is you now have a cloud data center operating the old skill set that isn’t helping your team to move along the path and just as quickly you can move to native Azure virtual machines and get the skill modernization along the way. 0:18:49.30 –> 0:18:56.470 Concurrency As part of that migration, so I would really suggest that you pick one or two which is modernize the workload along the way. 0:18:56.480 –> 0:18:58.360 Concurrency We’ll talk about that in a second. 0:18:58.890 –> 0:19:1.490 Ayman Mageed The other piece to it is as much as a vs. 0:19:1.500 –> 0:19:3.520 Ayman Mageed Sounds very alluring. 0:19:3.630 –> 0:19:17.100 Ayman Mageed There’s there’s a different service solution that you have to spin up to create that that Azure VMware service and it it requires some very unique networking. 0:19:17.150 –> 0:19:37.720 Ayman Mageed So when if you are operating at scale or are looking to scale up, whether it’s through acquisition or just natural growth of the organization, you’re gonna need to really architect that networking to make sure that, umm, you don’t have these these disconnects between other dependencies. 0:19:37.730 –> 0:19:48.960 Ayman Mageed The other piece of it is if you’re maintaining a hybrid environment, remember that you’re paying for every time that you’re either backing up outside of Azure or having those egress costs. 0:19:48.970 –> 0:20:10.620 Ayman Mageed So AVS creates a little bit more of a unique kind of snowflake, whereas if you’re trying to continue to scale up, it’s much easier to set a landing zone that we’re going to talk about here shortly, talk about landing zone and and be able to scale that and add and subtract at a moment’s notice, especially when we get into that. 0:20:10.680 –> 0:20:23.70 Ayman Mageed The DevOps piece of infrastructure as code and just being able to tear something down and build it right back from noon instead of kind of creating that unique environment essentially that we have to tiptoe around. 0:20:24.440 –> 0:20:24.880 Concurrency Were present. 0:20:27.350 –> 0:20:27.690 Concurrency OK. 0:20:30.310 –> 0:20:38.10 Concurrency So let’s move over to a little bit more detailed version of that Horizon 1, Horizon 2, Horizon 3 ecosystem. 0:20:38.20 –> 0:20:50.820 Concurrency So if you think about this as H1H2H3, OK, so each of these represent a service component that is gonna move from one domain to another, and I’m not going to like. 0:20:53.50 –> 0:21:0.820 Concurrency Could be single one of these, but as you stare at the slide, you can start to see the monetization that those different ecosystems. 0:21:0.890 –> 0:21:3.640 Concurrency So I’ll give you a kind of quick example here would be like. 0:21:3.650 –> 0:21:7.900 Concurrency Let’s say you have an on premise, RDS, Citrix or Horizon. 0:21:10.580 –> 0:21:14.310 Concurrency Sort of VM environment for your personal virtual machines. 0:21:15.140 –> 0:21:18.330 Concurrency I would not lift and shift your RDS servers in Azure. 0:21:18.840 –> 0:21:30.250 Concurrency It is as fast to migrate those RDS clusters into Azure Virtual Desktop and Oregon combination with Windows 365. 0:21:34.70 –> 0:21:47.490 Concurrency Because the Azure Virtual desktop essentially is RDS with an additional features like being able to run teams inside of the AVS instance and you don’t have to administer the infrastructure that sits underneath it. So. 0:21:50.820 –> 0:21:57.580 Concurrency Focht run that need to run really fast, like mergers and acquisitions that have an exit date with a TSA end date. 0:21:57.870 –> 0:22:1.410 Concurrency We still are moving straight to ads rather than lifting the RDS clusters. 0:22:2.640 –> 0:22:13.70 Concurrency On the other hand, if you are operating a Windows Server running on VM Ware, you’re probably like a non RDS server. 0:22:13.80 –> 0:22:29.480 Concurrency For example, you’re probably still moving that as is, but you’re changing it to a Windows Azure VM running on Azure with the future opportunity of moving to container based applications or functions as part of a refactor. 0:22:31.600 –> 0:22:33.230 Concurrency Umm, same kind of idea with SQL right? 0:22:33.240 –> 0:22:41.810 Concurrency Like if you’re running a Windows SQL Server, you could list that Windows, SQL Server, or you might consider flipping it over to like a managed instance or something along those lines. 0:22:42.480 –> 0:22:51.280 Concurrency It’s sometimes people overestimate the opportunity to refactor the SQL side of the picture or things like Azure SQL as part of the initial move. 0:22:53.950 –> 0:22:58.380 Concurrency But a near term opportunity to horizon three to say I still got all these Windows SQL servers. 0:22:58.650 –> 0:23:1.540 Concurrency Can I pivot these over to Azure SQL as part of the picture? 0:23:1.850 –> 0:23:7.0 Concurrency So understand that some of these shifts can happen really fast and some of these shifts cannot. 0:23:7.570 –> 0:23:24.150 Concurrency The other part of this picture that I just want you to consider, and we’ll talk a little bit more about it later because as you perform this move, there’s a whole variety of tools like backup, patching, monitoring that are gonna move to modern tools that are, Ajay, don’t feel tempted. 0:23:24.340 –> 0:23:30.770 Concurrency We will be tempted, but avoid the temptation to drag your management a state with you to the cloud. 0:23:31.520 –> 0:23:32.0 Concurrency Look at. 0:23:33.860 –> 0:23:56.510 Concurrency That a new so you can leverage tools that they come with the platform B can be orchestrated through infrastructure as code and leverage that modern skill set that lets you work at the Horizon 3 level leverage in Azure backup or a modern kind of duplicate of that Azure update management, Azure monitor get yourself using those native tools as part of that move. 0:23:56.520 –> 0:24:2.110 Concurrency And then you ultimately decommission huge parts of your on premises state that you’re not going to be running anymore. 0:24:3.0 –> 0:24:6.590 Concurrency So lots of opportunity here to think about how it moves from point A to point B. 0:24:6.900 –> 0:24:8.310 Ayman Mageed I guess, Nathan, I have a question. 0:24:8.860 –> 0:24:12.620 Ayman Mageed I am overwhelmed by the fact that I have to do all this right? 0:24:12.980 –> 0:24:24.250 Ayman Mageed Do I need to go into this horizon 3 refactored environment from scratch or what if my teams not upskilled to the point where we can potentially handle that? 0:24:24.260 –> 0:24:35.420 Ayman Mageed Yeah, I know there’s options, but could I potentially stay in the horizon two period for, you know, X amount of time until we can get to to that level? 0:24:35.770 –> 0:24:37.970 Ayman Mageed What does that work or how you see that work? 0:24:36.950 –> 0:24:37.940 Concurrency No, I agree. 0:24:38.350 –> 0:24:46.520 Concurrency I mean, Horizon 2 is the typical landing zone for, you know, mergers and acquisitions project or one someone trying to leave the data center. 0:24:46.850 –> 0:24:56.610 Concurrency I think the Horizon 3 is that incremental move that happens post the Horizon 2 Migration typically and you know it’s I I don’t. 0:24:56.620 –> 0:25:13.980 Concurrency I’m curious about your perspective like one of the things that I found is that that the biggest question mark and we hit this later on like cost is when do I apply reserved instances and when don’t I like, let’s say that I’ve I’ve moved my SQL servers in my Windows servers. 0:25:10.40 –> 0:25:10.570 Ayman Mageed Ah yes. 0:25:15.470 –> 0:25:24.200 Concurrency I know that if I do reserved instances, I gained tons of cost benefit, but I also don’t want to over reserve instance myself if I’m going to be scaling it back. 0:25:24.730 –> 0:25:35.380 Concurrency So there’s I think there’s a lift, a lift Dr efficiencies and then apply the zipped instances. 0:25:32.20 –> 0:25:32.430 Ayman Mageed That’s right. 0:25:34.500 –> 0:25:45.390 Ayman Mageed There’s there is, I mean, this probably could be a whole nother topic, but I think the the idea would be you have this band aid fix immediate moment where you reduce what you can decommission. 0:25:35.390 –> 0:25:36.340 Concurrency It’s similar to what you’ve been. 0:25:45.400 –> 0:26:0.60 Ayman Mageed We talked about and then with the remaining portion you have to commit to some sort of level of RIS because you’re going to you, you want to reduce the pricing, there is actually a break even calculator that allows you to. 0:26:1.710 –> 0:26:10.500 Ayman Mageed You know, if even if you picked A3 year reserved instance, there’s a point in that three years where you’ve made your money back and it’s not always the three year mark, right. 0:26:10.510 –> 0:26:15.290 Ayman Mageed So it’s often what I found is anywhere between the 14 to 16 months. 0:26:15.330 –> 0:26:22.350 Ayman Mageed So in in times for certain workloads or certain SKUs, it makes sense to just to pay for the three year. 0:26:22.360 –> 0:26:23.530 Ayman Mageed But you have to make this calculation. 0:26:23.540 –> 0:26:32.350 Ayman Mageed Obviously pay for the three year because you know that I for sure will have it for the 16 months break even or whatever the break even portion is. 0:26:32.660 –> 0:26:35.210 Ayman Mageed The other piece is where are we applying reserve instances? 0:26:35.220 –> 0:26:48.310 Ayman Mageed If we can standardize on a set of skews that majority of our engineers are going to use these types of virtual machines, then we can standardize that at the management group level rather than at specific subscriptions. 0:26:48.440 –> 0:26:58.90 Ayman Mageed And then as you add, subtract that skew, it doesn’t really make that much of a difference in the grand scheme of things, because there’s always that capacity to to manage. 0:26:58.330 –> 0:27:3.720 Ayman Mageed And so you definitely need to have some sort of quarterly finops type plan. 0:27:3.730 –> 0:27:6.660 Ayman Mageed At a minimum, I would say maybe quarterly. 0:27:6.750 –> 0:27:8.120 Ayman Mageed Ideally, it’s a weekly thing. 0:27:8.130 –> 0:27:9.460 Ayman Mageed That’s what we were used to do. 0:27:9.590 –> 0:27:13.900 Ayman Mageed And then you’re looking for anomalies that you can catch. 0:27:13.940 –> 0:27:31.440 Ayman Mageed And then there’s times where you say, OK, this is our operational plan, but then we have a queue X date where we’re going to reduce a certain footprint and this kind of opens up again to when we think of cloud as an enterprise solution and not just one. 0:27:31.490 –> 0:27:44.270 Ayman Mageed Technology is how are we integrating with things like our portfolio management, right and and tiering of our applications because that can help us decide what goes into that refactored environment section. 0:27:44.280 –> 0:28:7.470 Ayman Mageed Are we going for a Tier 1 application and going kind of risk it for the biscuit type situation or are we going to yeah, start with some of our Tier 3 Tier 4 applications because we have a little bit of room to learn there and and get some lessons learned and then move into those those more unique or critical cache cow type applications. 0:28:7.480 –> 0:28:15.740 Ayman Mageed So this is probably something that I’d love to kind of reexposure with everybody here, because that’s definitely another factor that we need to consider. 0:28:17.90 –> 0:28:18.460 Concurrency There percent 100%. 0:28:20.720 –> 0:28:34.10 Concurrency OK, so with all this understood I want to go through some recommendations that we have associated with as you start to plan this a state and there was a question that came up about all right, cool. 0:28:34.20 –> 0:28:36.130 Concurrency Like I want to consider. 0:28:37.990 –> 0:28:41.320 Concurrency What it looks like to exist in the cloud as a next step. 0:28:43.770 –> 0:28:46.820 Concurrency I think that’s a a variety of different elements we’re going to talk about. 0:28:46.830 –> 0:28:48.370 Concurrency One is what does the orange look like? 0:28:49.390 –> 0:28:50.780 Concurrency What are the skills look like? 0:28:50.950 –> 0:28:53.680 Concurrency What does the platform look like that I’m establishing? 0:28:53.750 –> 0:28:59.420 Concurrency But then the the big question everybody probably has is how do I like factor and understand the cost of that? 0:29:0.90 –> 0:29:4.670 Concurrency So recommendation like 4 or something, that’s like a top. 0:29:4.680 –> 0:29:9.40 Concurrency You see one of those clickbait things and it says #4 is the best. 0:29:9.390 –> 0:29:13.120 Concurrency That one is that topic is going to be in our recommendation. 0:29:13.130 –> 0:29:13.430 Concurrency So. 0:29:13.470 –> 0:29:18.560 Concurrency So a couple of topics coming, but I think that is going to be one that really addresses that question, Sir. 0:29:18.570 –> 0:29:19.780 Concurrency So we’ll get to that. 0:29:20.180 –> 0:29:25.10 Concurrency We’re gonna start it, though, by talking about like, as you consider this move. 0:29:26.550 –> 0:29:37.600 Concurrency You’re also planning for the future, like even if you said I’m maintaining my old data center, some people have said like I’m just leaving the old data center here and everything new goes to the cloud as refactored. 0:29:37.610 –> 0:29:38.40 Concurrency A state. 0:29:38.50 –> 0:29:45.460 Concurrency What does that organization look like and how do I prepare for even new workloads going there as I drain the old one? 0:29:45.470 –> 0:29:50.380 Concurrency I think this is one of the more critical steps because it’s about transforming your people, not just about the cost. 0:29:50.390 –> 0:29:51.700 Concurrency We’ll hit the cost thing in a second. 0:29:52.530 –> 0:30:8.800 Concurrency So the first thing I think you need to consider as you’re going down this path and I’m gonna move through a few of these slides relatively fast is your operational model and this is very much tied to how big you are like how how big, how complicated and what is the operational model of the business itself. 0:30:8.910 –> 0:30:33.150 Concurrency Most organizations are gonna fall into this category, which is this idea that you establish a cloud center of excellence that facilitates centralized control ecosystems where it is mainly the administrator of the app and the infrastructure and scenarios where it’s not where like you’re the IT is like the infrastructure admin or the is maybe even just the governor. 0:30:33.550 –> 0:30:37.40 Concurrency But then the area of the business is totally accountable for that app. 0:30:37.550 –> 0:30:58.290 Concurrency You’re establishing an operational model that facilitates you owning and managing that ecosystem and then on the other sides of this, you see things like decentralized operations, which is like startup territory and the distributed operations where the company is almost like a holding company and you have business one. 0:30:58.350 –> 0:31:3.710 Concurrency But then you have like other businesses that operate as their own satellites that have their own distributed ecosystems. 0:31:4.320 –> 0:31:9.820 Concurrency Most companies sort of fall into this spot right here, so I’m not gonna dream this. 0:31:9.930 –> 0:31:17.630 Concurrency I think it’s important to understand that you will have responsibilities and IT that need to be established in order to move to a cloud of state. 0:31:17.640 –> 0:31:24.860 Concurrency So if you’re like, I wanna be out of the VMware state on premise tomorrow, these are the three elements you’re gonna need to have in place in order to be successful. 0:31:24.870 –> 0:31:35.140 Concurrency First you need to have a part of your organization and understands what governance looks like for the cloud, and that means you’re establishing the policies and controls associated with that ecosystem. 0:31:35.670 –> 0:31:50.630 Concurrency You’re building a product project support capability that’s in in main extent, probably supporting your move like probably supporting the like the actual migration of the workloads from the on premise environment to the modern environment. 0:31:50.980 –> 0:32:5.250 Concurrency And then there’s an enablement function in organizations that aren’t even moving from their existing estate, and they’re just focusing on the newest state, are mainly doing one and three with two being the Migration Support. 0:32:5.520 –> 0:32:7.960 Concurrency One is governance, 3 is Enablement. 0:32:7.970 –> 0:32:15.640 Concurrency They’re always a balancing you can’t govern something that there isn’t a purpose for, so you’re not going to just govern. 0:32:15.650 –> 0:32:18.650 Concurrency You need to enable to, but you also can enable without governance. 0:32:18.660 –> 0:32:35.340 Concurrency You need both in the Enablement function is building that finops capability and combining that with onboarding new workloads in a responsible and appropriate way and helping those those assets to get established in a pattern that you can actively support. 0:32:35.660 –> 0:32:53.640 Concurrency And Finops is probably one of the most important components of what you need to get skills around as you move out because you were talking about that Ayman, like the extent to which you have your hands on the dials of reserved instances, licensing benefits optimization is real. 0:32:53.710 –> 0:33:2.900 Concurrency Like the difference between great, good and not good is many, many, many millions of dollars in larger estates. 0:33:3.570 –> 0:33:6.770 Concurrency So something to get you know, very, very good at is that. 0:33:11.420 –> 0:33:21.880 Ayman Mageed This is my pet project topic so I will I will refrain from taking us down that that aisle, but it’s definitely it. 0:33:18.410 –> 0:33:18.910 Concurrency Thank you. 0:33:21.890 –> 0:33:34.300 Ayman Mageed It is a beauty when you can get down to forecasting cloud costs ± 5 to 8% versus ± 15 to 30% because Finance starts to love you too. 0:33:34.670 –> 0:33:35.680 Ayman Mageed And when? 0:33:34.780 –> 0:33:35.340 Concurrency 100%. 0:33:35.690 –> 0:33:39.510 Ayman Mageed When you’re when you’re friends with finance and things get easier, I don’t know why. 0:33:43.270 –> 0:33:45.500 Concurrency OK, I’m gonna put this up on the screen. 0:33:45.510 –> 0:33:46.450 Concurrency You can screenshot it. 0:33:46.460 –> 0:34:9.80 Concurrency You can also get the deck from us later, but this is sort of a scaling path, so as you are thinking about the skills that your team needs to start to gain as they are moving into operating in Azure environment as opposed to an on premise environment, you’re gonna, I would suggest that you not just expect them to get training, but expect them to get certifications. 0:34:9.270 –> 0:34:14.500 Concurrency And I know sometimes we think like certifications, they’re not necessarily, they’re not representative of real life. 0:34:15.70 –> 0:34:17.280 Concurrency I I I think that’s starting to change. 0:34:17.490 –> 0:34:25.280 Concurrency And I and I think that they are representative of at least the base skills that are helpful for a person to understand as they’re operating these environments. 0:34:25.440 –> 0:34:29.740 Concurrency So as you look at the different lanes start to think about how you can move your people up those paths. 0:34:30.760 –> 0:34:39.530 Concurrency So in conjunction with this, your recommendation two is in addition to having the corporate governance, you also need to establish the management framework that’s going to exist. 0:34:40.0 –> 0:34:45.130 Concurrency So a lot of organizations, they start going into their new ecosystem. 0:34:45.240 –> 0:35:5.500 Concurrency They have like a subscription that was established and they start to consolidate everything underneath the few subscriptions and they find very frequently that that that to doesn’t build them a manageable cloud ecosystem that lets them take advantage of some of the scalable structures of the cloud, but also runs in the subscription boundaries. 0:35:5.630 –> 0:35:16.640 Concurrency So we recommend you not go toward models where you’re heavily consolidating under subscriptions and also don’t go towards Strategies: where you have so many subscriptions but no structure around them. 0:35:18.230 –> 0:35:22.140 Concurrency Build toward a structure that’s called the cloud, an option framework. 0:35:22.150 –> 0:35:23.840 Concurrency It’s very well documented at this point. 0:35:24.50 –> 0:35:40.560 Concurrency We’re we’re living in an age where many of the best practices have been pretty well laid out in which you can notice here is that the separations of concerns that exist in this starts to get to another fundamental why behind moving to cloud ecosystems. 0:35:40.570 –> 0:36:3.170 Concurrency It’s the opportunity to address separations of concerns associated with parts of the ecosystem, walling off identity in your domain controllers, walling off your management ecosystem components like your tooling, walling off the connectivity to the on premise environment, having a discrete sandbox that has no connectivity to the on premise environment or the operational environment. 0:36:3.580 –> 0:36:9.370 Concurrency And then building what’s called landing zones, which is this idea of a vertical stack of applications. 0:36:9.500 –> 0:36:14.210 Concurrency Each of these beams subscriptions that new workloads move into or mine. 0:36:15.480 –> 0:36:17.300 Concurrency There was moving too. 0:36:17.360 –> 0:36:22.280 Concurrency And what this is really facilitating is a scale of environment that is siloed from each other. 0:36:22.470 –> 0:36:33.30 Concurrency That allows you to be able to mitigate not just the cost questions, but also mitigate some of the exposure that you had on premise, especially things like ransomware and other exposure items. 0:36:33.990 –> 0:36:35.760 Concurrency So that’s going to come into play quite a bit. 0:36:36.10 –> 0:36:54.330 Concurrency We could talk for probably an hour just on CAFE alone, but understand that there is a best practice in terms of how to establish this and this becomes that box essentially becomes the center of gravity around all these other boxes that circle around it and represent the other pieces that sort of. 0:36:57.460 –> 0:37:3.830 Concurrency Tim, as a data workload, lands or app workload lands or you’re doing something in fabric. 0:37:3.840 –> 0:37:6.430 Concurrency All of this sort of attaches to that greater picture. 0:37:7.620 –> 0:37:12.700 Concurrency So you’ll also find that you start to establish patterns as you’re migrating into the environment. 0:37:12.710 –> 0:37:17.120 Concurrency This may be a lift and shift scenario, or it might be. 0:37:19.500 –> 0:37:24.70 Concurrency In public, internal or external facing app, you’re going to make some decisions around what am I? 0:37:24.130 –> 0:37:29.50 Concurrency What is the pattern that I’m that I’m adopted, so it might be like I’m losing private endpoints. 0:37:29.60 –> 0:37:34.900 Concurrency I’m using public endpoints and all this fits into different types of scenarios, so you’ll be establishing some path. 0:37:36.930 –> 0:37:46.230 Concurrency And the way it’s part of your governance estate, so maybe for internal facing app, external facing app or of kind of representative other alternative associated with that. 0:37:47.990 –> 0:37:57.860 Concurrency So this leads us into a conversation of as I’m moving out of my VMware environment into the cloud, how do I change my deployment practices? 0:38:0.120 –> 0:38:10.890 Concurrency In this represents one of the more important shifts that you’re going to see from skills development standpoint and why I don’t recommend maybe AVMS in comparison to deploying virtual machines. 0:38:11.600 –> 0:38:25.270 Concurrency And that’s primarily because we’re attempting to lay down an economy just to that’s maintainable later and we want an operational model that doesn’t leave the sort of haves and have nots of the sort of manageable ecosystem. 0:38:25.480 –> 0:38:31.980 Concurrency And the first thing to understand is that as you move into the cloud, some of the decisions you make in your provisioning process are going to affect downstream on the thin OPS side. 0:38:33.0 –> 0:38:37.420 Concurrency So you’re establishing this understanding of at least the. 0:38:40.100 –> 0:38:55.110 Concurrency Name and its relationship to the workloads that sit underneath it, and this may have a relationship also to your ITSM system over time in as you’re doing that provisioning, you’re getting the opportunity to be able to capture all these different data. 0:38:57.540 –> 0:39:18.860 Concurrency So as I take as I inventory workload from on premise, I relate that to the servers that I have, I’m capturing each of these elements and these elements get captured within the context of tags and classifications and documentation that I then used to apply policies to the migrated workloads such as like what kind of encryption do I need or how am I applying cost? 0:39:18.870 –> 0:39:21.790 Concurrency Or are there certain elements of the infrastructure missing? 0:39:21.800 –> 0:39:34.700 Concurrency If it’s highly available application, so these are captured along the way that that allows you to support is service provisioning workflow not just for the new workloads, but for everything you’ve moved as well. 0:39:34.750 –> 0:39:35.730 Concurrency And this is one of the. 0:39:38.160 –> 0:39:41.360 Concurrency I just point for a second in your on premise environment. 0:39:41.430 –> 0:39:48.160 Concurrency If you had to redeploy one of your workloads, even the virtual machines pretty frequently, we can’t do that when we move it into the cloud. 0:39:48.170 –> 0:39:54.130 Concurrency One of the things we do along with it is build a service provisioning workload for even the old workloads. 0:39:54.430 –> 0:39:58.700 Concurrency So like as you migrate that virtual machine, you should be able to redeploy that view. 0:39:59.100 –> 0:39:59.300 Concurrency Yeah. 0:40:0.700 –> 0:40:1.0 Concurrency Umm. 0:40:1.550 –> 0:40:11.880 Concurrency And then restore a backup to it based upon a save configuration you have that is redeployable and that isn’t that it’s relatively trivial to do because it’s already infrastructure as code in the cloud. 0:40:11.990 –> 0:40:14.470 Concurrency So we can just take advantage of that as a deployment ecosystem. 0:40:16.120 –> 0:40:29.950 Ayman Mageed And one of the coolest things as I was learning Cloud actually was create going through the motions of creating something on the console and realizing that I can actually pull the back end arm template that was used to create it. 0:40:30.10 –> 0:40:43.450 Ayman Mageed So it was a learning as I like on the job learning, essentially because I was able to go through it and say, oh, this is exactly what configurations are parameters it use and how how it went through it. 0:40:44.40 –> 0:40:52.650 Ayman Mageed I I don’t want to discredit or not discredit, but just kind of gloss over this piece and the importance that it brings when we talk about service provisioning. 0:40:52.940 –> 0:40:59.590 Ayman Mageed This opens up again the world of how do I bring other enterprise systems into play with what I’m doing in the cloud? 0:41:0.100 –> 0:41:4.970 Ayman Mageed I worked with a few clients where they were looking for a service catalog for the developers. 0:41:4.980 –> 0:41:7.470 Ayman Mageed We’re talking about innovation and speed to market here. 0:41:7.610 –> 0:41:16.700 Ayman Mageed When you’re in, when everybody’s a technology company, even Goodyear tires is a technology company because even though they’re making tires and automated facturing line, they have an application. 0:41:16.710 –> 0:41:18.460 Ayman Mageed They have web apps et cetera. 0:41:18.570 –> 0:41:21.120 Ayman Mageed And so those developers? 0:41:21.130 –> 0:41:38.800 Ayman Mageed Well, it wasn’t good your but that client was able to provide this front end called vending machine for their developers and they’re able to create ephemeral, you know, point in time virtual machines that they could use for testing, for development, for hosting their applications. 0:41:38.930 –> 0:41:49.270 Ayman Mageed And it was available to them because of the infrastructure as code that was set up and the ability to like going back to the tagging, being able to understand who owns it. 0:41:49.340 –> 0:42:0.330 Ayman Mageed What does that total cost of ownership for each application and Oregon Department, however, we want to set up, the organization gives us that more holistic picture of as a CIO. 0:42:0.710 –> 0:42:6.480 Ayman Mageed Can I tell my organization I know where every every dollar is spent? 0:42:6.490 –> 0:42:12.70 Ayman Mageed There might be some shared services costs we can talk about that on the side, but most of my dollars are spent on the cloud. 0:42:12.80 –> 0:42:15.300 Ayman Mageed I contribute to a cost center, a department or an application. 0:42:15.740 –> 0:42:34.710 Ayman Mageed I can also attribute it to production development or test or QA, so having that flexibility and setting those areas I mentioned in the comments section, setting those guardrails up first prevents a lot of headache at the back end because when we realize this midway through half the time, we usually will recommend. 0:42:35.20 –> 0:42:44.650 Ayman Mageed Let’s just blow it up and start over because it’s so much easier to start from a fresh foundational set of guardrails than to try to back back our way into it. 0:42:44.840 –> 0:43:1.710 Ayman Mageed So this is an area that is like I think all of us I concurrency are very passionate about how do we get to that level of innovation where we’re not just lifting and shifting but making business accelerators through through the cloud. 0:43:2.390 –> 0:43:4.520 Concurrency You hit a huge point on that. 0:43:4.570 –> 0:43:9.190 Concurrency Uh, that cost attributation attributation point. 0:43:10.410 –> 0:43:16.640 Concurrency It’s it truly is like a delineation between highly functional and less functional organizations. 0:43:17.90 –> 0:43:27.480 Concurrency The highly functional top quartile businesses are the ones that actually you transfer the costs to that business area and the ones that are less capable are the ones that kind of hold it all in it. 0:43:27.490 –> 0:43:46.770 Concurrency But just try to show it back that ability to actually move the cost in move also the ownership of accountability response, possibly an accountability alignment of that cost to the area of the business that that truly is actually responsible for what’s happening there. 0:43:46.980 –> 0:43:49.10 Concurrency Big difference in terms of actual impact. 0:43:51.240 –> 0:43:52.490 Concurrency Umm, OK. 0:43:52.560 –> 0:43:54.10 Concurrency So we talked a little bit about this. 0:43:54.20 –> 0:44:19.40 Concurrency One thing that we really recommend is that as you pivot to the cloud, even if you’re migrating on premise workloads, that we assume the usage of the portal experience as a configuration plane, it should exist to see what’s there, react to it, maybe have it as a playground, but as we deploy our production workloads, we have the opportunity, even if it’s just like infrastructure as code. 0:44:19.250 –> 0:44:27.730 Concurrency Like let’s say if a virtual machine running some box product like even that virtual machine can still be deployed as infrastructure as code via via pipeline. 0:44:27.940 –> 0:44:37.130 Concurrency So allows the team that you’re working with to be able to move their skills up to a higher capability level as part of lifting that workload into the cloud. 0:44:39.460 –> 0:44:39.980 Concurrency OK so. 0:44:42.410 –> 0:44:46.340 Concurrency There are a couple just scriptions we hit policy and tagging a little bit. 0:44:46.540 –> 0:44:53.980 Concurrency Understand that, like the conversation we just had about the service provisioning process directly leads into how you manage the environment. 0:44:54.130 –> 0:44:58.900 Concurrency So if you are capturing the service and that allows you to tag it that then. 0:45:0.940 –> 0:45:12.50 Concurrency Lets you build Azure policies that apply to those workloads that then lead to security outcomes, cost outcomes, optimization, skilling outcomes. 0:45:12.440 –> 0:45:25.550 Concurrency When people try to start here because the financial manager comes to them and says you’re spending what on Azure, they’re already but they can’t fix it because they didn’t start here. 0:45:25.560 –> 0:45:28.490 Concurrency They didn’t start at the actual service naming and tagging. 0:45:28.600 –> 0:45:37.800 Concurrency It’s really difficult to go fix that later, starting with the the First things first, leading to enabling us to be able to address outcomes. 0:45:39.290 –> 0:45:44.280 Concurrency So there’s some tagging values we’ll give you this deck later if you want to capture some of those best practices. 0:45:44.830 –> 0:45:46.460 Concurrency Also, some best practices there? 0:45:46.510 –> 0:45:46.840 Concurrency OK. 0:45:46.850 –> 0:45:51.580 Concurrency So the big question in the beginning was like, what do I do with enterprise cost models? 0:45:51.590 –> 0:45:53.100 Concurrency How do I think about this? 0:45:53.550 –> 0:45:54.780 Concurrency This is this conversation. 0:45:55.30 –> 0:45:58.940 Concurrency So this is a little bit of like a shot of how I like to think about this. 0:45:58.950 –> 0:46:0.820 Concurrency So this is a tool that we use. 0:46:3.130 –> 0:46:7.620 Concurrency Umm to think about on premise costs and reserve costs and so on. 0:46:7.860 –> 0:46:11.140 Concurrency There’s two main dials that you don’t really three main dials that you have. 0:46:11.920 –> 0:46:16.820 Concurrency They’re the first dial is the scale back or eliminate workloads. 0:46:16.830 –> 0:46:17.150 Concurrency Dial. 0:46:17.280 –> 0:46:17.710 Concurrency OK. 0:46:18.220 –> 0:46:26.950 Concurrency So just optimization right sizing Dial that’s not tied to licensing, it’s just tied the like scaling things back and you always find that workloads. 0:46:26.960 –> 0:46:27.750 Concurrency Can be scaled back. 0:46:28.180 –> 0:46:41.560 Concurrency The second thing that you find is that you can apply hybrid use benefits, reserved instances, other kinds of sort of SAS orpas based reserve assets that now exist to be able to drive that cost down. 0:46:41.960 –> 0:46:51.70 Concurrency So when a lot of people look at the costs in the cloud, they take their as is cost as pay go and they’re or even into reserves and they’re like, ooh, like, there’s still a lot of money. 0:46:51.240 –> 0:46:59.160 Concurrency And yeah, it is, but your replacement cost I find is like some replacement cost on premises usually like somewhere in the here range. 0:47:1.860 –> 0:47:2.910 Concurrency It’s like it exists. 0:47:2.920 –> 0:47:3.900 Concurrency It’s it’s more. 0:47:3.940 –> 0:47:6.590 Concurrency It’s less expensive than pay as you go, but more expensive than reserves. 0:47:6.670 –> 0:47:7.880 Concurrency It’s like sits in that middle spot. 0:47:9.420 –> 0:47:11.140 Concurrency I find that this is the right direction. 0:47:12.10 –> 0:47:13.220 Concurrency You go from pays. 0:47:13.230 –> 0:47:24.600 Concurrency You go to right sizing to reserves and you’re you’re sort of moving down that slope trying to get the reserves quickly but also not reserving for things you’re not keeping. 0:47:24.870 –> 0:47:35.50 Concurrency And one of the things you find quickly is once you start using reserves on compute, it’s the storage costs that really are the main thing. 0:47:35.260 –> 0:47:37.350 Concurrency And then storage costs are different. 0:47:37.360 –> 0:47:42.950 Concurrency Part of the picture where I can start to eat into that by using things like archival costs that are represented here. 0:47:43.80 –> 0:47:57.970 Concurrency So I’ve got a customer that has like all sorts of expensive on premise storage hosting images that of different types of customer assets that go back ten years, but they’re using Tier 1 storage for virtually everything. 0:47:58.320 –> 0:48:12.830 Concurrency So moving them to the cloud to something like an AB Azure files moving the majority of that over to archival storage drastically changes even this storage picture because it starts to take them in a direction where the storage costs are really advantageous. 0:48:13.220 –> 0:48:15.210 Concurrency So you’ve got to look at it like this. 0:48:15.210 –> 0:48:21.230 Concurrency You’ve got to look at it in a picture of different dials that you’re working at to try to minimize as you move into the cloud. 0:48:21.240 –> 0:48:22.410 Concurrency Don’t just look at all. 0:48:22.480 –> 0:48:26.350 Concurrency If I take this versus AVS and I dump it into AVS, I’m gonna get this cost. 0:48:26.360 –> 0:48:29.790 Concurrency You want to look at in terms of how I can actually drive toward these optimizations. 0:48:31.120 –> 0:48:39.110 Concurrency And then when you look at the patterns of migration, you start to group those into waves and pause periods. 0:48:39.360 –> 0:49:12.80 Concurrency So this is just sort of an example what something like that like might look like if you take a look at those costs and you start to group applications into theoretical wave patterns and then you understand like in those waves, what is the cost associated with different parts of it and how have I pictured which I’m rehosting, which I’m rehosting in migrating and then which of those you I’m actually moving to SAS along the way like I you know in a certain sense like sass or modernize might be like I’m moving RDS over to an avd. 0:49:12.180 –> 0:49:20.830 Concurrency Instance, which is gonna be less expensive than shifting the actual service itself, but you have to build out a migration plan that helps you add a project. 0:49:20.840 –> 0:49:28.370 Concurrency The costs be and know how and modernizing along the way and see to know like how am I managing risk and impact along this path. 0:49:28.380 –> 0:49:45.880 Concurrency And you can see that in this project plan like we’ve got a lot of these kind of initial servers and then I’ve got SAP kind of at the tail end A cause it’s you’re building skills as you go and be it minimizes risk, but also it’s the biggest, it’s the biggest skillset chunk that has to get moved along with it. 0:49:48.290 –> 0:49:49.280 Concurrency Email take a breath. 0:49:49.330 –> 0:49:49.790 Concurrency Any thoughts? 0:49:51.10 –> 0:50:5.80 Ayman Mageed So I’m catching up on all the comments here, but from what I from what I heard I I agree with that statement I I I think to also bring this all into reality, cause usually all of this requires a business case. 0:49:53.680 –> 0:49:54.140 Concurrency Well, there you go. 0:50:5.130 –> 0:50:8.780 Ayman Mageed Is the risk appetite right? 0:50:8.790 –> 0:50:9.800 Ayman Mageed How are we? 0:50:9.870 –> 0:50:16.440 Ayman Mageed Are we deciding to move things that are business critical first or are we trying to move things that are less critical? 0:50:16.530 –> 0:50:17.550 Ayman Mageed How are we? 0:50:19.510 –> 0:50:26.970 Ayman Mageed Presenting that to our counterparts in the in the way of saying what are the benefits and outcomes that we’re going to achieve by doing this right? 0:50:27.20 –> 0:50:44.830 Ayman Mageed I mentioned here on the side application modernization lifecycle too, right where it’s OK in the interim to have certain decisions when we’re migrating, but we should have in mind an idea of I’m doing this because in X amount of years we’re going to be refactoring this application. 0:50:27.170 –> 0:50:27.430 Concurrency Umm. 0:50:44.840 –> 0:51:0.800 Ayman Mageed We’re going to be doing XY and Z because where I’ve seen some organizations kind of misstep is we’re gonna sweat this application until it dies and then it’ll be a problem to fix. 0:51:0.860 –> 0:51:12.860 Ayman Mageed The early someone else’s problem to fix it when we have the technology and the opportunities ahead of us, I almost joke that sometimes the best place to go work is a startup because they don’t have any of this tech debt, right? 0:51:12.960 –> 0:51:15.50 Ayman Mageed We’re starting brand new from the cloud. 0:51:15.60 –> 0:51:16.670 Ayman Mageed I don’t have to worry about my mainframes. 0:51:16.680 –> 0:51:19.30 Ayman Mageed I don’t have to worry about some of this older technology. 0:51:19.200 –> 0:51:30.610 Ayman Mageed Windows 2003 servers and I can just start working on innovation, but those are the companies that blow up and are our competitors because we are not taking care of our going back to the garage, right? 0:51:30.620 –> 0:51:33.520 Ayman Mageed We’re not taking care of our garage and so. 0:51:33.530 –> 0:51:49.540 Ayman Mageed So I think that’s that’s kind of where I will always bring it back to the business minded concept of sure we have all this tech debt, but I’d rather focus on some of the tech debt that’s gonna Support improving and and improving our bottom line for the business. 0:51:49.750 –> 0:51:53.750 Ayman Mageed Then some of the tech that that like, yes, it needs to happen, but we could probably do this. 0:51:55.170 –> 0:51:59.780 Ayman Mageed Kinda drag ourselves along and do this as we can as a point of opportunity. 0:52:0.560 –> 0:52:2.50 Concurrency Each point news point. 0:52:2.790 –> 0:52:11.970 Concurrency So as we bounce from that into finops, I think we’ve been talking about this as we go, but finops really is a representation of what you did to get to this point. 0:52:12.300 –> 0:52:18.450 Concurrency If you did the right pre work in your migration, you did the right pre work of understanding the services and grouping the applications. 0:52:18.660 –> 0:52:22.490 Concurrency You’re gonna have an effective thin OPS operation on the tail end of. 0:52:22.500 –> 0:52:23.690 Concurrency You didn’t do that pre work. 0:52:23.750 –> 0:52:26.150 Concurrency You’re gonna struggle within OPS because you’re gonna see costs. 0:52:26.160 –> 0:52:31.600 Concurrency You’re gonna see names of resources, but you’re gonna struggle to attribute those resources to the right targets. 0:52:31.970 –> 0:52:39.420 Concurrency So understanding like what they are and what they cost and who’s causing that cost to establish, is one of the most critical elements of the overall ecosystem. 0:52:39.810 –> 0:52:51.920 Concurrency So thinking about that in terms of billing, Routing, understanding how to build a true center of excellence associated with grouping those resources is a critical part of building your migrated estate. 0:52:51.930 –> 0:53:5.740 Concurrency So you can hold yourself accountable to costs that you’ve already promised to your executive team as a result of your pre planning, and that what I found is that building a finops team really happens in two stages. 0:53:6.90 –> 0:53:12.480 Concurrency So as you are establishing that competency and you’ve done the right pre work, there’s two elements to maintain and build. 0:53:12.560 –> 0:53:18.120 Concurrency The first is you’re building a finops team that has a goal, and the Finops team. 0:53:18.430 –> 0:53:19.930 Concurrency When I say team, I mean this. 0:53:21.40 –> 0:53:26.930 Concurrency We all know that there’s opportunity that exists in our our cloud environments to maintain and manage costs. 0:53:27.520 –> 0:53:37.110 Concurrency I usually tell people that 20 to 40% of their cloud environments waste and that’s people leaving things on, not using reserved instances, not scaling resources back. 0:53:37.320 –> 0:53:40.230 Concurrency There’s a tremendous amount of opportunity there, but we don’t get it. 0:53:40.240 –> 0:53:40.540 Concurrency Why? 0:53:40.550 –> 0:53:41.150 Concurrency Why don’t we get it? 0:53:41.160 –> 0:53:49.330 Concurrency It’s because we don’t apply time and expertise to that problem, so it sits there and it keeps billing us and it’s like taxes. 0:53:49.340 –> 0:53:58.980 Concurrency If I knew if I actually knew how much I was writing in that check to the government every single month, I’d be a little bit more upset than the fact that it just comes out of my paycheck. 0:53:58.990 –> 0:54:3.570 Concurrency And I really look at it until I do my taxes over here, it’s kind of that way of thin up sometimes. 0:54:3.580 –> 0:54:6.440 Concurrency Like we just know it’s getting billed to us, but we don’t put the energy in. 0:54:6.850 –> 0:54:30.180 Concurrency So Finops is really about assigning time and energy to that cost reduction and holding ourselves accountable and knowing that for every hour I apply, if I can pay off that hour double in finops cost reduction, I have created value for the organization and it’s moving from sort of phase one establishment to phase two reality of driving those costs down. 0:54:31.200 –> 0:54:42.430 Concurrency So ultimately, the best teams they’re building and managing a backlog around those lanes on the organization that we talked about, one of which is finops, the other is establishing governance rules. 0:54:42.440 –> 0:54:53.320 Concurrency And then those two in the middle of Enablement and product support, so how that moves into where we go from here is fourfold O, how do you get started? 0:54:53.370 –> 0:54:59.780 Concurrency Now if you I don’t leaving this the end and this is like maybe one of the main topics, but it’s all right. 0:54:59.790 –> 0:55:0.140 Concurrency Cool. 0:55:0.150 –> 0:55:1.340 Concurrency It’s all sounds interesting. 0:55:1.350 –> 0:55:2.740 Concurrency Like I wanna learn more. 0:55:2.750 –> 0:55:4.400 Concurrency I wanna take the next step. 0:55:4.730 –> 0:55:6.980 Concurrency The first step is let’s run an assessment. 0:55:7.130 –> 0:55:8.820 Concurrency Let’s look at what you have today. 0:55:8.910 –> 0:55:10.660 Concurrency Let’s use some tooling and gather it. 0:55:10.670 –> 0:55:13.100 Concurrency Let’s understand what the costs are in different environments. 0:55:13.110 –> 0:55:20.850 Concurrency Let’s look at the common refactoring options that are realistic to do quickly and understand that picture. 0:55:20.860 –> 0:55:26.230 Concurrency So you can cost justify and that then lets you set the table for all right? 0:55:26.240 –> 0:55:32.280 Concurrency I do need to have enterprise governance in place and maybe you already have some workloads in the cloud and you haven’t built enterprise governance. 0:55:32.290 –> 0:55:33.330 Concurrency Let’s catch that up. 0:55:33.640 –> 0:55:44.900 Concurrency But perhaps you haven’t really put a lot out there yet, and This is why you’re going to then do that governance establishment, which then lets you position new refactored workloads as you maybe didn’t even migrate it. 0:55:44.910 –> 0:55:55.270 Concurrency You’re just enabling new things like AVD or or a new deployed modern at the state or a data workload or an AI workload that like even if I had the on premises state, I still needed to do. 0:55:55.800 –> 0:56:3.20 Concurrency But then I might then move into migrating the age data center because I now have run the Azure assessment. 0:56:3.180 –> 0:56:5.500 Concurrency I know what the actual costs are going to look like. 0:56:5.510 –> 0:56:18.570 Concurrency I know what it means when I refactor them or I scale them or I apply reserved instances and I know that this is better than me renewing my on premise data warehouse data contract on premise contract, my Colo contract and I’m going to take that action. 0:56:18.800 –> 0:56:19.790 Concurrency But we gotta start. 0:56:19.800 –> 0:56:20.790 Concurrency We gotta start here. 0:56:20.860 –> 0:56:22.110 Concurrency Let’s have that conversation. 0:56:22.440 –> 0:56:25.870 Concurrency So we will do that on as an investment from concurrency. 0:56:25.880 –> 0:56:28.870 Concurrency We want to have that conversation with you. 0:56:27.170 –> 0:56:27.460 Ayman Mageed Umm. 0:56:28.960 –> 0:56:32.210 Concurrency We will leverage tooling to be able to help run that Azure assessment. 0:56:32.220 –> 0:56:40.260 Concurrency Microsoft is oftentimes an asset to us to be able to run some of those, but we bring our own flavor to it to help you understand what that picture looks like in the future. 0:56:40.270 –> 0:56:41.270 Concurrency And that’s in the survey. 0:56:41.280 –> 0:56:42.290 Concurrency So as you leave today. 0:56:44.500 –> 0:56:47.990 Concurrency Talk about how we can help you and provide some feedback as to things. 0:56:48.0 –> 0:56:48.770 Concurrency Maybe we would have. 0:56:48.790 –> 0:56:53.370 Concurrency You would have liked to see that maybe we didn’t tell you or if you love the session, we love to know that too. 0:56:53.440 –> 0:57:0.290 Concurrency We always want to improve what we’re doing and providing information because part of what our goal is is to lift the whole community around knowledge. 0:57:0.440 –> 0:57:3.660 Concurrency But then also be an asset to you as you try to take it forward. So. 0:57:3.550 –> 0:57:16.580 Ayman Mageed 100% One Piece, one piece that I just thought of actually as as I was looking at this is there’s a regulatory compliance piece to this as well or as we’re making these assessments as we’re figuring out, do we need to get out of the data center, do I need to do XY and Z? 0:57:16.590 –> 0:57:17.940 Ayman Mageed How do I set up my landing zone? 0:57:18.230 –> 0:57:29.220 Ayman Mageed But there are so many blueprints that we have available to us to look at ISO 27,001 or, you know, NIST and or even PCI DSS. 0:57:29.230 –> 0:57:29.400 Ayman Mageed Right. 0:57:29.410 –> 0:57:38.730 Ayman Mageed Like depending on where you are in your industry, there’s probably something regulatory body that is causing you some obviously some headache, but some necessity. 0:57:38.840 –> 0:57:41.870 Ayman Mageed But then there’s also the piece of how do I partner with security? 0:57:41.880 –> 0:57:51.450 Ayman Mageed Or maybe I am the security like I security kind of falls under my my wing as well and I want to reduce the threat vector that I have in my environment right? 0:57:51.460 –> 0:57:55.930 Ayman Mageed Like if I’m on VMware Azure insert all these other technologies. 0:57:56.60 –> 0:58:10.230 Ayman Mageed There’s so many vulnerabilities that are coming out there and we know that the threat actors are much better and faster at finding them than we are because we have all these other operational tasks where they, you know, they’re they’re just nudging away at a computer. 0:58:10.300 –> 0:58:24.200 Ayman Mageed So so this this is another area to consider when we’re looking at running an Azure assessment is what are the functional requirements for my organization and my cloud so that we can make it fit for purpose as well. 0:58:29.990 –> 0:58:30.440 Concurrency Awesome. 0:58:31.290 –> 0:58:34.660 Concurrency OK, we hope you have a great rest of the day. 0:58:34.720 –> 0:58:36.580 Concurrency We’re so thankful that you spent some time with us. 0:58:36.590 –> 0:58:47.180 Concurrency We’re looking forward to having filed conversations and hopefully we’ll see you on the flip side, this will be available as a recording and then you’re also able to get the deck later if it would be an asset for you. 0:58:47.190 –> 0:58:50.50 Concurrency So feel free to ask for that and we’d be happy to give you the deck as well. 0:58:51.430 –> 0:58:51.990 Concurrency Have a great day. 0:58:52.580 –> 0:58:53.170 Ayman Mageed Thank you everyone.