To SQL or noSQL, Why is that the Question with Chris Harris

• couchbase.com: https://couchbase.com
• LinkedIn: https://www.linkedin.com/in/chris-harris-5451953/
• Twitter: https://twitter.com/cj_harris5
  
  

Transcript

Announcer: Hello, and welcome to Screaming in the Cloud with your host, Chief Cloud Economist at The Duckbill Group, Corey Quinn. This weekly show features conversations with people doing interesting work in the world of cloud, thoughtful commentary on the state of the technical world, and ridiculous titles for which Corey refuses to apologize. This is Screaming in the Cloud.

Corey: Couchbase Capella Database-as-a-Service is flexible, full-featured and fully managed with built in access via key-value, SQL, and full-text search. Flexible JSON documents aligned to your applications and workloads. Build faster with blazing fast in-memory performance and automated replication and scaling while reducing cost. Capella has the best price performance of any fully managed document database. Visit couchbase.com/screaminginthecloud to try Capella today for free and be up and running in three minutes with no credit card required. Couchbase Capella: make your data sing.

Corey: This episode is sponsored in part by LaunchDarkly. Take a look at what it takes to get your code into production. I’m going to just guess that it’s awful because it’s always awful. No one loves their deployment process. What if launching new features didn’t require you to do a full-on code and possibly infrastructure deploy? What if you could test on a small subset of users and then roll it back immediately if results aren’t what you expect? LaunchDarkly does exactly this. To learn more, visit launchdarkly.com and tell them Corey sent you, and watch for the wince.

Corey: Welcome to Screaming in the Cloud. I’m Corey Quinn. One of the stranger parts of running this show is when I have a promoted guest episode like this one, where someone comes on, and great, “Oh, where do you work?” And the answer is a database company. Well, great, unless it’s Route 53, it’s clearly not the best database in the world, but let’s talk about how you’re making a strong showing for number two.

It sounds like it’s this whole ridiculous, negging nonsense or whatever the kids are calling it these days, but that’s not how it's intended. Today’s promoted guest is Chris Harris, who’s the Vice President of Global Field Engineering at Couchbase. Chris, thank you for joining me and I really hope I got it right and that Couchbase is a database company or that makes no sense whatsoever.

Chris: It’s great to be on the show, and thank you for the invitation. I’m looking forward to it. Yeah, we’re a database company. That’s exactly what we do.

Corey: I always find it interesting when companies start pivoting from a thing that they were and, “What do you do?” “We build databases.” [unintelligible 00:01:29] getting out of that space it’s, “What do you do?” “We’re a finance company.” And then there’s a period of time in which they start reframing what they do. It’s, “We’re a data platform.” Or, “We’re now a tech company.”

Really? Because I don’t get that sense in any meaningful perspective. Couchbase was founded as a database company. You went public last year—congratulations on that—and now you continue to say, “Yes, we’re a database company,” rather than an everything trying to eat the world all at the same time, mostly ineffectively, company. So, what kind of database are you folks?

Chris: So, if you look at the database world, you can see—I’ve been in the space for quite some time now, a good few years, and I’ve had the privilege, if you like, of being at other database companies, been in the analytics space, and I’m here at Couchbase. But if you look at the history over the last—let’s just not go back all the way that far, but let’s go back to, like, ten years ago, everybody was building their applications on traditional relational databases. And what you saw is that the Oracle and MySQL, as traditional databases of the world. And then… probably at the time, we realized that with, talking ten years ago where we had this demand for high throughput of data, next generation of applications will be built, and then people realized the traditional database architectures weren’t going to cut it, if you like. And it spawned this industry.

You know, a big NoSQL market was created. And you have document databases, and then you have graph databases, and then you have analytics databases, and you have search databases, and then you have every sort of database you could possibly think of type database that’s out there in the world.

Corey: You have so many kinds you need to keep track of it all inside of the database.

Chris: That’s what you have to do, right? [laugh]. But the interesting thing is it became different types of database. And even see this in many of the code providers today, right, that you have multiple different types of databases no matter what you’re trying to do, right? So, we kind of went—Couchbase kind of took a step back and went, okay, we were originally a cache, right, this is where we came from, and then kind of built that into a document database, and then kind of went to the market and went hold on here, rather than it being let’s call it a noSQL versus SQL discussion, why can’t it just be a database, right?

Why can’t you have a SQLite interface on top of a modern architecture? Why can you do that, right? Why can’t you have the flexibility and architectural [unintelligible 00:04:16] of a JSON-based database with the interface of—with SQL, and then analytics built on top of that, right? So, why can’t you have the power of SQL on the next generation architecture? So, that’s kind of where we fit in the world.

Corey: When we talk about origin stories and where things come from, well, let’s start with you. I guess the impolite version of the question is, “Why on earth would you be in a space like this for so long?” But you’ve been on a lot of interesting places doing somewhat similar things. You were at Cloudera, you were at Hortonworks until you apparently heard a who or whatnot, you were at MongoDB, you were at VMware, you were at Red Hat. And that’s going reverse chronologically, but it’s clear that you’re very focused on a particular expression of a particular problem. Why are you the way that you are? Only pretend that’s a polite question.

Chris: “Why am I the way that I am?” Well, first of all, I love technology, right? That’s the key. And I think many of us in the industry would definitely say that, right? I started off in core engineering, building—I know some people today wouldn’t probably remember this, but when you had Chip and Pin where your credit card and you have to type it in and put in a pin number, that was created originally in the UK, and then went off and built e-commerce websites for retailers.

Well, that then turned into—was a common theme that I kept seeing is that lot of the technology that we’re using was open-source technology. And that kind of got me into the open-source movement, if you like, and I was lucky enough to then join Red Hat when they built middleware frameworks, so got into that space there. And then did a lot of innovation in the middleware space. Went to SpringSource and we did some great work there in the Java Development Framework space. But what became interesting is that—you still see it today—like, in this innovation happening in that middleware space and there’s some great innovation happening, right?

There’s all this stuff with Lambda and serverless architecture that’s out there, but they always came back to, we’ve got the database, this thing that is in the architecture if it goes down, you’re stuffed, right? This is where the core value of your company is sitting. So, then that got me interested to see what innovation was happening in this space. And as I say, I got into this field in the early stages of NoSQL, where there was that spawn of new database technologies being created. And then from there, it was like, “Okay, let’s get into what was happening in the analytic space.”

Again, I’m still in the Hortonworks, and Cloudera space, that’s all open-source. But it came down to this is different types of databases that were required different types of skills. And then I started talking to the team here, who was like, “How can we take as great innovation and leverage the skills I already have?” And I thought that was an interesting point.

Corey: In the interest of full disclosure, I tend to take the exact inverse approach to the way that you did. When I was going through the worlds of systems administrator, than rebadged as DevOps, or SRE, or systems engineer, production engineer, whatever we’re calling ourselves this week, I was always focused primarily on stateless things like web servers, or whatnot because it turns out that—this should be no surprise to longtime listeners of this show—but I’m really bad with computers. And most other things, too; I just brute force my way through it. And that’s hilarious when you keep taking down web servers you can push a button and recreate. When you do that to a database or anything that’s stateful, it leaves a mark.

And if you do it the wrong way, just well enough, you might not have a company anymore, so your DR plan starts to look a lot more like updating your resume. So, I always tried to shy away from things that played to my specific weaknesses that would, you know, follow me around like a stink. You, on the other hand, apparently sound—how to frame it—you know, good at things, and in a way that I never was. So you’re—ah, you see a problem, you’re running towards it trying to help fix it; I’m trying to how do I keep myself away from making the problem worse is my first approach. It seems like you have definitely been focused on not just data themselves—I mean at some level, [if it was a 00:08:55] pure data problem, it feels like we’d be talking a lot more about storage, but rather how to wind up organizing that data, how to wind up presenting that data, and the relationship that data has to other things that are going on. I’m not speaking in the sense of a traditional relational database, necessarily, but the idea of how that data empowers businesses and enables them to do different things. Is that directionally a fair synopsis of how you see it?

Chris: I think the [unintelligible 00:09:21] thing is what I would agree with. What makes it really interesting to me is what we enable people to do with that data, and being able to build, kind of, really fascinating innovation applications that are affecting their underlying businesses, right, from it could be health care, it could be airlines, financial services, some really high, interesting use cases that people are doing that are leveraging the database to be able to drive that level of innovation. Because it’s very difficult; I can build some sophisticated application, but if I can’t get the performance out of my database, I have a pretty poor experience to my users in today’s world. Because, fortunately or unfortunately, people aren’t very patient, right? If you have a website that doesn’t return very quickly, a customer’s gone like minutes ago. You literally got to instantly respond to someone. That’s a challenging problem.

Corey: It absolutely is. Something that I found as I’ve talked to a bunch of different companies operating in different ways is the requirements on data stores are generally very different depending upon primarily latency and performance. There’s only so long people that are going to watch the spinning circle of doom on a website spin before they realize they’re going to go somewhere that has its act together. Conversely, for a lot of business intelligence and analytics queries, there are an awful lot of stories where the thing that people care about is that we actually have to have the results of this query by noon on Thursday. And there are very different use cases for that, and some companies seem to be focused very much on, “We’re going to solve both of those use cases extremes and everything in between with the same product offering,” and others tend to say, “Okay, this is the area of the market we’re going to focus on.”

You could also say that this is an expression of the larger industry question of do I want, more or less a one-size-fits-most database that’s general-purpose, or do I want very specific purpose-built databases based upon the use case and the problem? Where do you find yourself on that spectrum?

Chris: I find myself on that spectrum is that there’s—if you want to describe it at a high level and we can break it down, there’s operational-type databases, where I’d say Couchbase fits where you’re talking about, I’ve just built an application; I’m talking to the live user, right, this is what I care about, and when I’m talking about speed and performance here, I’m talking about something that returned within milliseconds of response time. I’m playing an online game, or I’m doing online betting on a sports game. That has to be pretty much instant, right? If we’re playing multiplayer games and you’re doing something, then I want to be able to see what you’re doing straight away, right? People don’t expect it to lay there.

If you’re looking at streaming—people do this with Couchbase—streaming the Olympic Games or Super Bowl in the US, and you want to be able to be there, that whole profile management of that user has to be instant, have that stream to you has to be instant. People use telephone calls and use Couchbase to do, behind the scenes, profile management, right, so they know who you are who’s making that call. That’s an operational database problem. That’s not a traditional analytical problem, right? So, there’s a whole other space in the database world for analytics, right, which is bringing all the data together into one place, and I’ll help you do data science, AI, machine learning, be able to crunch and compute large volumes of data. If I get back to you, rather than a week in an hour, that’s great, but that’s not operational. That’s analytical.

Corey: In data center environments, it’s an argument to be made for going in a bunch of different directions; we’re going to use a bunch of different data stores to store all these things. Because, generally speaking, the marginal cost of moving data from one of your data storage systems to another one of your data storage systems, one rack row over is fairly small, whereas in cloud, effectively, there are no real capacity constraints anymore until you can get the bill, but that’s the entire problem where a lot of the transfer for these things is metered per gigabyte. So, there’s a increased desire on a lot of architectural pressures, to wind up making sure that where the data lives, it stays. And whatever it is that you do with that data, it should be able to operate on that data in a way that fits your performance characteristic requirements in the place that it currently is. And on the one hand, I can definitely see that driving a lot of decisions people have made.

The counterargument is that it feels a little weird when the cost constraints of how the cloud providers—mostly you, AWS—have decided to build these things out. And that, in turn, is shaping your entire approach to not just your architecture, but your systems design of how data winds up working its way through your lifecycle. It’s frustrating, on some level, especially given that they themselves offer something like 15 distinct managed databases offerings but more announced all the time. It becomes very difficult not only to disambiguate between all of them but to afford moving data from one to the next.

Chris: The affordability is an interesting discussion, right, because you can look at it from a billing perspective and go absolutely, there’s a challenge associated to that. Then is a question of where is my data because it’s spread across all these different services; that’s another challenge. And then you have the challenge of, okay, the cost associated to having developers build applications against all these different types of services because they all require different APIs and different ways of programming. So that’s, there’s a cost associated everywhere.

Corey: Oh, by far and away, the most expensive part of your AWS, or any cloud spend, is not the infrastructure itself; it’s the payroll expense associated with the people working on it. People always cost more than infrastructure. If not, something very strange is going on.

Chris: But then you look at it, and you go, okay, if that’s the case, I kind of use the analogy, right, that it’s like a car, where everyone is talking these days about the electric car [that’s going 00:16:05] on that path, right? Now, I should be able—if I was getting an electric car—think of it now, I actually have one—that I can get in the car and I can drive it like any other car. I know what a steering wheel is, I know where my pedals work, it looks and feels like a normal car. But architecturally it’s fundamentally different how it operates. So, why can you apply that same thing, that same analogy to a database, right?

So, why can’t I have the ability from an operational perspective, [unintelligible 00:16:42] talk about operational databases, not necessarily, I don’t know, full-blown analytical databases, but operationally being able to say I can store the data in an enterprise database; I can use that to leverage my SQL skills like I have before, and also use it to have a document store under operational analytics, to eventing, to full-stack search, key things that people want to do operationally, but keeping the data together in one database, like an iPhone. I want a database to have these capabilities; I don’t want to have all these different types of devices that are everywhere. I want, you know, my iPhone to be able to go to have the capabilities that I’m using. Or my car, to feel like I’m driving a car; doesn’t matter if the underlying architecture of the engine changed. That’s great, I want the benefit, but I want to be able to drive it in the same way that I’ve driven any other car out there. And that’s kind of trying to solve multiple problems that because you’re trying to solve the issue of skills.

Corey: It’s one of the hard challenges out there, and I think your car analogy can even be extended a bit further because in the early days of the automobile, you were more or less taking some significant risk by driving a car if you weren’t also mechanically inclined and to fix it yourself. And in time, we’ve sort of seen that continue to evolve where they mostly work, and now they work really reliably. And then you take it even a step beyond that, and all right now I’m just going to pay a car service so someone else has to deal with the car and a driver, and I don’t have to deal with any of that aspect. And it feels like there are certain parallels, similar to that, toward the end of last year, 2021, you folks, more or less moved away from you can have it in any color you want, as long as you run it yourself—more or less—into offering a fully-managed database-as-a-service cloud option called Capella, which, on the ads for this show, I periodically sing because if you didn’t want me to do that, you would not have named it Capella. Now, what was it that inspired you folks to say, “Hm, we could actually offer this as a managed service ourselves?”

It’s definitely a direction a lot of companies have gone in, but usually, they have to wait to be forced into it by—let’s be serious for a second here—Amazon launching the Amazon Basics version of whatever it is themselves and, “Okay, well, they validated our market for us. Let’s explore it.”

Chris: If you look at that, you go Couchbase has been around for a good few years now selling, as you point out, high-performance databases to large-scale enterprises, on real mission-critical, people call it tier-zero type applications, high-performance applications. And these are some of the most fascinating, most innovative type of applications that I’ve been involved with through my career. Now, how can we take that capability, provide it to the mass market if you’d like, to be able to give it to people that don’t need to have a large number of people out there managing their own infrastructure, being able to understand how to finely tune that underlying infrastructure to get the level of performance that you need from high-performance databases. Now, there are use cases for doing that, so it’s not one or the other. It’s not that you have to go all-in.

There are particular companies out there that, for the economics reasons, for the use case reasons that are running today on-premise, and there’s a rational reason for why they do that, right? But for a lot of people out there, whether they’re leveraging the cloud, there’s an opportunity here to take the power of the database, allow us to then manage it for people, take away that complexity of it, but being able to give them the power so they can leverage their skills, take advantage of Couchbase far easier than ever have been able to in the past. It’s opened up a bigger market for us, to summarize your question.

Corey: This episode is sponsored by our friends at Oracle Cloud. Counting the pennies, but still dreaming of deploying apps instead of “Hello, World” demos? Allow me to introduce you to Oracle’s Always Free tier. It provides over 20 free services and infrastructure, networking, databases, observability, management, and security. And—let me be clear here—it’s actually free. There’s no surprise billing until you intentionally and proactively upgrade your account. This means you can provision a virtual machine instance or spin up an autonomous database that manages itself, all while gaining the networking, load balancing, and storage resources that somehow never quite make it into most free tiers needed to support the application that you want to build. With Always Free, you can do things like run small-scale applications or do proof-of-concept testing without spending a dime. You know that I always like to put asterisks next to the word free? This is actually free, no asterisk. Start now. Visit snark.cloud/oci-free that’s snark.cloud/oci-free.

Corey: One way that I tend to evaluate where a given vendor sees themselves—and it’s sort of an odd thing to do, but given that I do fix AWS bills for a living, it probably makes sense—I wind up pulling up the website, I ignore the baseline stuff of the, “This is what Gartner says,” and here’s a giant series of scrolls. I just go for the hamburger menu and I look for, “All right, where’s the pricing information?” Because pricing speaks a lot. And there are two things I generally try to find. One is, is there a free trial that I can basically click and get started working with?

Because invariably, I’m trying to beat my head off of a problem at two in the morning, and if it’s, “Oh, talk to a salesperson,” well as a hobbyist, or as an engineer who does not have signing authority for things, but it’s talk to sales, I realize, “Oh, yeah. One, I probably can’t afford it. Two, it’s going to be a week or so before I can actually make progress on this, and I’m hoping to get something up by sunrise, and it’s probably not for me.” Conversely, the enterprise tier should always have a, “Call for details,” because that is a signal to large enterprise procurement departments and buyers and the rest were it’s, “Oh, we will never accept default terms. We always want them customized. We also don’t believe in signing any contract without at least two commas in it.”

Great. So, being able to speak to both ends of the market is one of those critical things that you folks absolutely nail that. What I like is the fact that if someone has a problem that they’re experimenting with at two in the morning, they can get started with your database-as-a-service platform—Capella; or however you want to sing it—and they don’t need to wind up talking to you folks directly, first. There’s no long-term commitments, there’s no [unintelligible 00:22:39] of the infrastructure themselves. There’s no getting hounded for the rest of their days over making a purchase for something that didn’t pan out.

To me, that’s always been the real innovation and breakthrough of cloud is that I can spend a few hours some evening kicking around an idea, and if it doesn’t work, I can turn it off and spend 17 cents on the process, whereas if it does work, I can keep scaling up without at some point having to replace all of the Raspberry Pi’s and popsicle sticks, I build things with real enterprise-grade stuff. There’s a real accessibility and democratization that is entered into it. So, I’m always excited when I see companies that are embracing that model. Because, yeah, I’m a grumpy old sysadmin because it’s not like there’s a second kind of sysadmin, but—and I have a particular exposure and experience level with these things that I can’t expect modern developers to work on. They have an idea, they want to launch something, and they just need a database to throw things against and put data into, and ideally get it back again when they query later. And that empowers them to move forward.

They’re not in this because they really want to run virtual machines themselves and get those set up and secured and patched and hardened, and then install the software on top of it, and, “Why is it not working? Oh, security groups, how you vex me again. I’ll just open you to the entire world,” and so on. And we know where that path leads. So, it’s nice to see that there is an accessible option there.

Conversely, if you come at this with an approach if we are only available in our hosted cloud environment, well now those big enterprise companies that have, you know, compliance concerns are going to have some thoughts for you, none of them particularly pleasant in some cases. So, I like the fact that you’re able to expand your offering to encompass different user personas without also, I don’t know, turning what has historically been a database into now it’s an LDAP server, and trying to eat the world, piece by piece, component by component.

Chris: It’s interesting that you say that because I think there’s a number of things that you’re touching on that were to me, if you look at us as a company in particularly this space, there’s a lot of focus around the community and the open-source community. And I think there’s an element of how do you make it accessible to people as a community as a whole? And then you kind of go down the path of, “Okay, let’s allow people,”—as a developer, let’s think of it this way, right, the ultimate thing they want to do, and you touched upon it there, is they want to build an application. They get passionate about building the application or maybe even in the weekend, and they got this funky idea that they’re going to literally knock some code out.

And I remember my fond memories of being an application engineer of being able to sit down for hours just been able to put my ideas into code and watch it execute. The last thing that I want to do is get to the point where I get the database and go, oh, here we go. This is going to take me a bunch of hours, now, and I’m going to set it all up and do other stuff. And I almost literally want to be able to click a few buttons—

Corey: You know what I want to do tonight? Feel really dumb as I tackle a problem I don’t fully understand. Gr—I’d love smacking into walls that point out my own ignorance. It’s discouraging as hell. I’m right there with you.

Chris: Yeah, you don’t want to do that, right? So, you almost want to make like the database disappear for people, right? You want to be able to just say, like, “Here’s your command. Off you go. Bring the data back. Bring it back in full. Allow it to scale.”

Because you want that developer to have that experience of not breaking their flow. And what do you want them to be able to be so excited about the application and innovation that they’ve built, that they want to go and show that teammates? They want to say, “Look at the great thing I built over the weekend. Look at this, this is amazing.” Right?

And then be able to get all the teammates pretty excited about what they built in a way in which they can try it out really easily, right? They can take this little thing that they built into the database, click some buttons, and off we go, right? And now your development team is super excited about some of the great innovation that you have. But you also have to have the reverse. You have to have the architecturally sound, so then when you get to the architect, if you like, who is looking at the bigger picture of what’s the future going to look like? Is it the right technology? Is this something that we can bring into the organization? And you know, this is a cool bit of application you just built me, but you know, is this realistic that I can deploy this thing?

And this is where you start going back into it still has to have high performance, the security has to be there, the scalability has to be there so that I can potentially—I can start small and grow this thing horizontally as I see the requirements coming. There are different set of requirements architecturally, so we’re looking at—you know, as a company, our key focus is how do you drive that developer community so that you give the people the freedom to build the next generation of applications in the simplest way [unintelligible 00:27:35], say with free trials, click some buttons, have the database up in minutes, but also then being able to have that capability in the underlying database to take it to the architect. That’s what our core focus is every day.

Corey: I agree with everything that you’re saying. You’re making an awful lot of great points, but for me, the proof in the pudding is the second thing that I tend to look at on your website after the pricing page, and that is your list of customers. Because it’s always interesting when someone talks about how they’re revolutionizing everything, and this is the way to go, and everyone who’s anyone is doing these things. And then you look at their customer page and either they don’t have one, which is telling, or the customers on that page are terrifying in that, “Wow, that sounds like a whole bunch of fly-by-night startups whose primary industry is scamming people.”

You have a bunch of household blue-chip names as well as a bunch of newer companies that are very clearly not what people think of as legacy—you know, that condescending engineering term that means it makes money. It’s across the board, it is broad-spectrum, and it is companies that absolutely know exactly what it is that they’re doing when it comes to these things. That to me is far more convincing than almost anything else that can be said because it’s—look, you can come on and talk to me about anything you want about your product, and I can dismiss it and, “Yeah, whatever. Great.” But when I start talking to customers, as I did prior to recording this episode, and seeing how they talk about you folks, that to me is what reaffirms that, okay, this is actually something that has legs and is solving real customer problems.

Because early stage, it’s, “We have this idea for this company we’re going to build that it’s going to be great.” “Awesome. Go talk to more customers.” That is a default, safe piece of advice generically you can give to anyone. And it’s easy to give and hard to take.

I’ve been saying this for years, and I still screwed it up and we started trying to launch a SaaS product here called DuckTools. Yeah, it turns out that we didn’t talk to enough customers first about what they’re actually trying to achieve, and we assumed we knew the answers. It’s an easy mistake to make. What I really appreciate is—about a Couchbase in particular—is not just the fact that you have all of these customer references, but the fact that each one talks about what the value to the business is not just in terms of, “Oh yes, now we can query data and there was no way for us to do that before.” Of course, people have found ways to do that since business started.

Instead, it’s much more about this is how it made it more efficient, more optimal, how it unlocked possibilities and capabilities for us. That alone tells me that there definitely is significant value that you’re delivering to customers. In my own business, whenever I think I’ve seen it all, I have to do is talk to one more customer and learn something new. What have you seen in recent memory, from a customer, that surprised you about how they’re using Couchbase?

Chris: You look at that, and you can see—I could probably talk for hours on different types of customers, but it’s the ones that you can literally see in your life and you can reflect to, right? So, if you taken one of the biggest airlines that are out there today, they’re completely changing, kind of, the whole experience. And our whole experience of and how do I get feedback? Because Couchbase’s customers, [unintelligible 00:31:01] customer, right, is what they’re thinking about, right? They’re an airline.

So, these passengers; fine. But how many times have you got on a plane, and you see all these people, literally, there’s obviously the passengers, and then there’s the cabin crew, and then there’s the people on the ground, and then there’s the pilot, and for the sake of the discussion, the staff that are there are literally passing paper back and forth to each other. And surely there a better way to do this. And for someone who likes to solve complex technical problems, you go, “Wow, this is going to be a bit of a challenge.” Because if you want to collect feedback from an aeroplane in the air, [laugh] right, and you want to connect that to the ground data that people are having in terms of maintenance data, you want to do that across the world, in multiple different time zones, that’s pretty tricky problem to try to go solve, right?

So therefore, how do you get a database that is able to work remotely and on what people would call the edge; let’s just call it in this case in a device that’s literally a cabin crew member is carrying around with them that’s not connected because there is no connection because I’m in the middle of the air. But I want to pair it with the other cabin crew members that are around, right, in flight, and then when I land, I want to sync that data backup to the maintenance people. So, you need a database that’s able to operate on a device with no connection, and then being able to synchronize backup to a cloud database that is then collecting data from all the other flights around the world.

Corey: Synchronization sounds super easy until you actually try and do it, and then, “Oh, wow.” It’s like, you could cut to pieces by the edge cases.

Chris: And then people go, “Well, there’s no problem. There’s internet everywhere these days.” Yeah, sure there is. [laugh]. You get disconnected all of the time.

Corey: Not to name names. This is very evocative, an earlier episode of this show I had with Tyler Slove, who’s a senior manager over at United Airlines, about specifically how they’re approaching a lot of their own reimagining and the rest. It’s a fascinating use case, and as someone who’s a bit of a travel geek himself—you know, in the before times—that’s always an area of intense interest because it’s… I’m sorry, I’m still a little boy at heart; it’s magic to me. You get on a plane, you go somewhere else, close the doors, it opens it up, and you’re on the other side of the world. And now there’s internet on it? Oh, my God, who would have imagined such a thing?

Chris: Uh-huh. But that’s changing the experience for people. It’s just really fascinating.

Corey: Completely. And it’s empowering and unlocking that experience you’re talking about of being able to sync between the crews, about handling all this stuff behind the scenes. Everyone loves to complain about airlines because no one knows really how to run the massive logistical part of an airline. But the WiFi was a little bit slow or the food was cold; well, that’s something I know how to complain about Twitter.

Chris: [laugh].

Corey: It becomes this idea of almost a bikeshed problem expression, where it’s, “Oh, yeah. I’m just going to complain about things I can wrap my head around.” Yeah.

Chris: I was talking to somebody recently, and they were—swapping topics a little bit—and they were like, so—they were talking about innovation on some new web application that they built. And I literally have to explain them, and I said, “Well, if you think of it, the underlying whole technology stack that’s behind this for high-scale e-commerce, it’s sophisticated, right, because people will literally walk away from a page, an application, a mobile app, if they don’t get an instant response time. And that request has to literally travel, physically, quite a fair amount of distance, talk to multiple different types of technology, answer to that question, then come back to you instantly.” The sheer amount of technology that’s involved here of moving that data around is a complicated architectural problem to fix. A database only plays a small part of that. You can’t be the slowest player in the party.

Corey: No. And that is always the challenge is that when you’re looking at different use cases, there’s always a constraint, and how that constraint winds up manifesting in different ways, if it’s not the thing that’s slowing things down, it’s also not where the attention goes. If you have a single thing like, the database for example, slowing things down, everyone cares about improving databases, people focusing on, “Well, we’re going to improve the JavaScript load time on the website,” that’s not the problem. Find the bottleneck and focus on it. And although I’m generally a fan of picking a database and using that as a general-purpose thing until it makes sense not too—much like I am cloud providers—

[audio break 00:35:54]

Corey: —journey personally, where’s the best place to find you?

Chris: Clearly, if you want to find more about Couchbase, you can obviously go to couchbase.com. You kindly pointed out you can go and look at the trial for Capella and try out the tech. You’re more than welcome to do that as a free trial.

If you want to contact me particularly, you find me on LinkedIn; I’m Chris Harris at Couchbase. You’ll find me [unintelligible 00:36:26] with Chris Harris in general and probably find lots of them. In the UK, Chris Harris is a famous racing driver. That’s not me; it’s someone else. So, find me on LinkedIn, I’m sure it won’t be that difficult to find what you find. Or you can find me on Twitter.

Corey: And we will of course, but links to all of that into the [show notes 00:36:43]. I really want to thank you for being so generous with your time today. It’s always appreciated to talk to people who actually know what they’re doing.

Chris: You’re more than welcome. It’s been great to be on the show. Thanks, Corey.

Corey: Chris Harris, Vice President of Global Field Engineering at Couchbase. I’m Cloud Economist Corey Quinn, and this is Screaming in the Cloud. If you’ve enjoyed this podcast, please leave a five-star review on your podcast platform of choice, whereas if you’ve hated this podcast, please leave a five-star review on your podcast platform of choice along with an angry comment. I’m going to wind up using all of those angry comments, at one point, as a database.

Corey: If your AWS bill keeps rising and your blood pressure is doing the same, then you need The Duckbill Group. We help companies fix their AWS bill by making it smaller and less horrifying. The Duckbill Group works for you, not AWS. We tailor recommendations to your business and we get to the point. Visit duckbillgroup.com to get started.

Announcer: This has been a HumblePod production. Stay humble.