 OK, so my name is Marco, and I work at VMware on Tanzi. My job is I'm the product line manager for Tanzi Edge. I'm on Craig McLucky's team, and he talked last year. I hope you watched that video. It was a lot of fun. I wanted to kind of pick up where he left off. This is the only slide that has the word VMware in it, it's just his quote. And wanted to spend a couple of minutes reviewing what he was talking about. And since sometimes we were very early in our journey, maybe give you a little bit of an update and kind of share with you some of the themes that we're trying to resolve in a way that I think will hopefully be useful to you all in that if you're in this room, you are my friend because we're all trying to figure out Edge. Some of you may be very early in your journey. Some of you may be a little farther along. Hopefully, this is cool. So I wanted two key themes. Good. The font is too small for all of us to read. That was kind of on purpose. I don't want you to really read the screen. The slides will be uploaded later. But I thought what would be helpful is to expand on what Craig was talking about and kind of explain what our challenges are for helping figure out how to work on the Edge for relatively large companies. And the two things that I really wanted to drill into your mind today is if you are able to do these two things, I think you will find your journey to Edge to be a lot easier. And the first theme is to think of Edge as a cloud. We are really trying to think of when we deploy it to all infrastructures that we would like to get to the point where a developer can push software, an application, or other software to the Edge the same way that they're pushing to public cloud. It's taken us, what, 10, 12 years to figure public cloud out. So the most important thing, and I'm not going to read every line on the slide, is make sure to treat your Edge sites like they're ephemeral. Prepare for disaster in the same way that Amazon taught us, like, don't rely on that VM. Don't expect it to be there tomorrow morning. If you lose it, just be able to create a new one. Try to do the same thing with your Edge sites. So an interesting thing that comes up a lot when we talk with IT people who are trying to figure out how to instrument their stores is, how do I back up and restore my Edge site? Do you really want to do that? Don't back up and restore Edge. If you lose it, just throw it away and start over again. Did you keep persistent data there? Maybe you shouldn't keep persistent data at the Edge where somebody can spill a Coke on top of your machine. Find another way. And similarly, the second theme is use abstractions to push apps. I don't know about you, but I really enjoy using Puppet, Chef, Terraform, a bunch of these tools. And I feel pretty confident now that I can use any variety of systems to push my app pretty productively to one site, like the primary site. Maybe I'm even good at pushing to DR and I'm good at switching back and forth. I don't know about you, but I don't want to push to 1,000 sites if I have 1,000 stores or 10,000 sites. If you use an application platform, then it will be a lot easier for you to implement something like GitOps. And by the way, I'm really uncomfortable with the term GitOps. I feel like it's not very descriptive for what we're trying to do. You want to use a CD system that it was aware of, a source code control system that knows how to push that. It gets too many words. But if you can find a better term than GitOps, please share with me. I would like to know. So what are some of the challenges that you will find for these things? So one of the big ones that I want to call out today is when you're developing edge sites, it's really worthwhile to get people to stop thinking about we need five nines. All of us have just spent the last 10 years, 12 years, getting out of the data center into public cloud. We are very, very good at Kubernetes, especially running in highly available mode. My own documentation, I looked at my own product one day. It says, if you're running in highly available mode, you're not supported, period. It's like, well, hang on. That can't be right. If we take the store offline while it's closed, we actually have an eight hour maintenance window every day. So just try and push on yourself. Can I make it so that these sites are disposable so that I don't need to depend on them 100% of the time, but that if I do need high availability, I know how to invest on that appropriately so that my spend and my value are kind of manageable. Because anything you do will be multiplied by the number of sites you're at. So this isn't maybe the best place to put three rack mount servers underneath everybody's coffee machine. And similarly, for the application platform, what we're really trying to do is trying to figure out what does edge native mean. I think we've all gotten pretty comfortable with the idea of cloud native, but what is an edge native app? What does it mean to be edge native? Maybe it means that it's durable to poor latency. Maybe it's what we call effective for what we call submarine mode. Sometimes a submarine goes under the ship. Sorry, under the water. It's offline for a long time. Craig called this periscope mode, which I like to. Maybe it has a very thin connection. We've all gotten pretty used to being in the cloud where we have infinite bandwidth and it's always reliable. That's not going to be that way now. A lot of people are converting to public consumer-grade internet as opposed to team one lines, frame relays, Mac. That's not there, so we're still working on that. And by the way, you'll notice that there's only one word in the whole presentation that is highlighted and actually highlighted it twice, which is security. Everything else in my talk, you could pretty much discard. If we all get security right, then maybe we'll do a better job of succeeding at the edge. And then super briefly, what are the kind of key technologies that I think are worthwhile here? The highlights that I would like to cover on the technologies, like I'm actually really appreciative. Thank you so much for bringing up the exhaustive list, because I made sure not to put any names of technologies in mind, so we compliment each other well. Thank you. But I think one of the things that we're discovering is that, for edge, almost everybody has existing workloads. And I've been very careful not to insult anybody. Don't call them legacy workloads. These are the workloads that are running the phone system. They're running the lights. They're running the cash registers. That's not legacy. Those things need to continue running. So one of the big priorities we're finding when we go out and talk with people, like, what do you need? How do you want to be successful at the edge? That's fine. But don't take down the phone system. The cash registers don't need to run. So whatever we're doing, we're layering in. And I'm super excited about the Kubernetes API. I think it's conveniently, perfectly wonderful to deploy 1,000 times. But don't disrupt the VMs. Make sure that they're still able to run. Make sure that they still have their storage. Make sure that they can migrate if there's a hardware failure. And so I feel like that's something that's super important when thinking about running at the edges. Like, don't push the other stuff aside. Don't try and replace it. And I would also say, under key technologies related to the application platform, and if you want to talk about this afterwards, I would really love to hear it, is we're discovering that while everybody feels like they have a really good idea about how to deploy the API, how to get Kubernetes up and running, how to do all these things, then I say, yeah, but how are you going to distribute your container images? And everybody hits a big pause. It's a challenge, right? You can't just have a container registry in the cloud and expect everything to reach out to the cloud. Sometimes you're not going to be able to do that, or sometimes you won't have the bandwidth. How will you know when it's finished replicating? This is turning out to be a pretty considerable challenge. So it's part of our journey. I wanted to share that with you. I also don't want to keep you from the rest of the talks. So myself, Mayank, will be outside at the table when you want to chat. And if these are interesting ideas to you, or if you're curious to learn more about how we're choosing to solve them, like I didn't put any prescriptive stuff in this deck, please come and meet us. All right? Thank you. Thank you.