 Good and dog, bienvenue. Howdy, everybody. My name's Adi Folger. I'm a principal architect for Redis Labs. If you're so inclined, this is how you can get in touch with me or find out more about me. But really, I'm here today to talk about Redis, Redis Labs, Redis Enterprise, and Cloud Foundry, specifically Pivotal Cloud Foundry, and how they all work together. So without further ado, let's go ahead and get started. So just quick show of hands. How many of you have heard of Redis before? Cool, like everyone. How many of you use it in production? Cool, pretty much most of you. All right, so I'm not going to spend too much time talking about what Redis Open Source is. I think probably most of you, as just found out, are already familiar with that. But I am going to talk a lot about what we do at Redis Labs to help make people who love using Redis enjoy that experience even more. So Redis Labs is the company behind Redis Open Source. And essentially, these are the three main things that why people decide to use our technology. So from a performance perspective, as you already know, we blow everyone else out of the water. These benchmarks are actually a little bit dated. At this point, we're up to 50 million operations a second, being able to leverage a lot of different pieces of technology to be able to get there. And then, of course, Redis' memory is a service. You have all of these different data structures that allow you to build out very, very complex technology. And this is why we've been voted the most loved database on Stack Overflow now for two years, because it's so easy to use, so flexible. And then with Redis Labs, you can scale that up in a pretty significant way, as I just mentioned. The really cool thing that many of you probably are not aware of is that we now have a module API. Using that whole memory as a service metaphor that I mentioned before, this is essentially classes. So it allows you to write your own data structures. And doing so, we've taken Redis from having just nine or 10 kind of these core primitives to having over 50 in about a period of about a year and a half or so. So that includes JSON, search, and graph. JSON is, of course, kind of throw a serialized blob into Redis, and it will decompose it down into the various underlying data structures that make sense, and then allow you to be able to interact with it on a more granular level than you could traditionally. And you don't necessarily have to go through the overhead of deserializing and serializing that you used to have to do. Search is five times faster than currently what we're seeing with Elasticsearch. Full-text search with querying, grouping, all kinds of cool little tools. It's building out its own really robust query language, and we're seeing some really cool stuff happening with that in terms of customer usage. And then lastly, graph is based on graphblast. It's storing data in a really, really unique way. And so we're getting better performance than, say, Neo4j in a lot of circumstances. This powers all kinds of different things. Because it's this very simple toolkit, you can really use it for practically anything. And that's what many, many of our customers are adopting it, to be more and more of an operational database, an analytics platform, and figuring out how to bridge the gap between big data and operational data. So that kind of leads me into Cloud Foundry. We've worked very closely with Pivotal over the last several years to build out a really robust partnership. And then we were able to work together to provide an even better solution for high-scale enterprises. So what that comes down to is the fact that we're gonna be able to do replication out of the box for you. It's just checking a thing. It's not necessarily having to go in and mess with config files, like you have to with the open source. Our architecture lives on every single node, so there's no notion or single point of failure, which is really important when we start talking about high-scale data applications. And then with the underlying Bosch infrastructure, we're able to do zero downtime upgrades. And then of course, on the security side, role-based access control, we can do replication over SSL, which we have many financial services customers. That's a very important thing for them. And this is something that you can do out of the box versus something that you can't even do with open source today. And then we have a really strong multi-tenancy model that allows you to have multiple databases sharing on the same VM, versus something like Elastic Cache, which is doing the exact opposite. It's making you use as much machine time as possible to on as little resources as possible. On the replication side, we have a few different options for you. Your standard active passive hot standby for HA purposes. Our failovers generally take anything from less than a second to up to maybe about 30 seconds, depending on the severity of the failure. And then we can also do read replicas that allow you to be able to scale out reads or individual keys, which is very difficult to do, especially if you have a couple of hot keys that are getting most of your traffic. And then the really cool thing is this new active-active technology that's based on conflict resolution data types. And those, if you're not familiar with that, essentially a few mathematicians got drunk one night and were like, how do we solve data replication? Really deep and interesting topic. And they figured out how to, using various algorithms, come up with a deterministic way of dealing with conflict resolution across geographies that didn't necessarily have to have an atomic clock in every single data center, which is the case for Spanner and a lot of other technologies out there. If you're interested in that, my colleague Vic gave a talk yesterday. So if you can watch that video, he goes into a lot more detail on how that works. It's inside a PCF. We have Service Broker API. We're able to go in and create databases on the fly using either the REST API or the CFCLI API. And then we make that active-active technology very simple to configure and use. So this is a super, super high level of what that looks like. If you wanna get a little bit deeper, it looks something like this. So we're using that TCP router component as the backbone for this. And it's just something that you go in and you configure in ops manager. So you just go in and set up all of the different parameters in there. And then as well, you go in and you tell it all, where all these different clusters live by giving them that appropriate TCP router endpoint. And that sticks with the PCF networking paradigms, which is really important, right? So you build this really big box. The last thing that you wanna do is start to have to poke holes in it just to make your technology work. So the fact that we can follow those paradigms makes it even more secure. As well, you can go in and configure as many different plans as you want, some of which can be geo-replicated, some of which can just be active-passive, or just singletons, depending on what the use case is. And in the near-term future, we're also gonna be able to support multiple clusters with various different permutations of plans running in the same PCF foundation, which would allow you to be able to provide isolation between business units and various other things like that. It's available out there on PivNet now. So give it a download, try it out. Like I said, we have TCP router in there. It's the SysHog integration and all kinds of various bells and whistles to make your life easier. And yeah, quick. Any questions? Nope, thanks, y'all, for your time.