Screaming in the Cloud
aws-section-divider
Audio Icon
MongoDB’s Purposeful Application Data Platform with Sahir Azam
Episode Summary
For the first in-person episode in quite some time, Corey is joined by Sahir Azam, Chief Product Officer at MongoDB. Recording during the madness known as re:Invent, Sahir has graciously taken some time to bring us up to speed on what the folks at MongoDB have been up to, both in their work with AWS and outside of it. With re:Invent in the rearview, Sahir reflects upon MongoDB’s week well spent. Sahir and Corey discuss MongoDB’s history as a “staple of the internet” and how MongoDB is now set in stone as one of the leading cloud database companies. MongoDB has made purpose a central focus for its database build. This purpose is something the company provides to its customer base as a wide array of organizations make the transition to the cloud. Sahir and Corey discuss the nuances of general purpose databases and their potential shortfalls, plus how to craft a more specific database that allows space for innovation.
Episode Show Notes and Transcript
About Sahir

Sahir is responsible for product strategy across the MongoDB portfolio. He joined MongoDB in 2016 as SVP, Cloud Products & GTM to lead MongoDB’s cloud products and go-to-market strategy ahead of the launch of Atlas and helped grow the cloud business from zero to over $150 million annually. Sahir joined MongoDB from Sumo Logic, an SaaS machine-data analytics company, where he managed platform, pricing, packaging and technology partnerships. Before Sumo Logic, Sahir was the Director of Cloud Management Strategy & Evangelism at VMware, where he launched VMware’s first organically developed SaaS management product and helped grow the management tools business to over $1B in revenue. Earlier in his career, Sahir held a variety of technical and sales-focused roles at DynamicOps, BMC Software, and BladeLogic.


Links:
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: This episode is sponsored in part by our friends at Redis, the company behind the incredibly popular open source database that is not the bind DNS server. If you’re tired of managing open source Redis on your own, or you’re using one of the vanilla cloud caching services, these folks have you covered with the go to manage Redis service for global caching and primary database capabilities; Redis Enterprise. Set up a meeting with a Redis expert during re:Invent, and you’ll not only learn how you can become a Redis hero, but also have a chance to win some fun and exciting prizes. To learn more and deploy not only a cache but a single operational data platform for one Redis experience, visit redis.com/hero. Thats r-e-d-i-s.com/hero. And my thanks to my friends at Redis for sponsoring my ridiculous nonsense.  



Corey: Are you building cloud applications with a distributed team? Check out Teleport, an open source identity-aware access proxy for cloud resources. Teleport provides secure access to anything running somewhere behind NAT: SSH servers, Kubernetes clusters, internal web apps and databases. Teleport gives engineers superpowers!


  • Get access to everything via single sign-on with multi-factor.
  • List and see all SSH servers, kubernetes clusters or databases available to you.
  • Get instant access to them all using tools you already have.


Teleport ensures best security practices like role-based access, preventing data exfiltration, providing visibility and ensuring compliance. And best of all, Teleport is open source and a pleasure to use.


Download Teleport at https://goteleport.com. That’s goteleport.com



Corey: Welcome to Screaming in the Cloud. I’m Corey Quinn. For the first in-person recording in ages we have a promoted guest joining us from MongoDB. Sahir Azam is the chief product officer. Thank you so much for joining me.



Sahir: Thank you for having me, Corey. It’s really exciting to be able to talk and actually meet in person.



Corey: I know it feels a little scandalous these days, when we’re in a position of meeting people in person, like you’re almost like you’re doing something wrong somehow. So, MongoDB has been a staple of the internet for a long time. It’s oh good; another database to keep track of. What do you do these days? What is MongoDB in this ecosystem?



Sahir: That’s a great question. I think we’re fortunate that MongoDB has been very popular for a very long time. We’re seeing, you know, massive adoption grow across the globe and the massive developer community is sort of adopting the technology. What I would bring across is today MongoDB is really one of the leading cloud database companies in the world. The majority of the company’s business comes from our cloud service; we partner very heavily with AWS and other cloud providers on making sure we have global availability of that. That’s our flagship product.



And we’ve invested really heavily in the last I would say, five or six years, and really extending the capabilities of the product to not just be the, sort of, database for modern web scale applications, but also to be able to handle mission-critical use cases across every vertical, you know, enterprises to startups, and doing so in a way that really empowers a general purpose strategy for any app they want to build.



Corey: You’re talking about general purpose which, I guess, leads to the obvious question that AWS has been pushing for a while, the idea of purpose-built databases, which makes sense from a certain point of view, and then they, of course, take that way beyond the bounds of normalcy. I don’t know what the job is for someone whose role is to disambiguate between the 20 different databases that they offer by, who knows, probably the end of this year. And I don’t know what that looks like. What’s your take on that whole idea of a different database for every problem slash every customer slash every employee slash API request.



Sahir: What we see is customers clearly moved to the cloud because they want to be able to move faster, innovate faster, be more competitive in whatever market or business or organization they’re in. And certainly, I think the days of a single vendor database to rule all use cases are gone. We’re not [laugh] by any means supportive of that. However, the idea that you would have 15 different databases that need to be rationalized, integrated, scripted together, frankly, may be interesting for technical teams who want to cobble together, you know, a bespoke architecture. But when we look at it from, sort of a skills, repeatability, cost, simplicity, perspective of architecture, we’re seeing these, sort of like, almost like Rube Goldbergian sort of architectures.



And in a large organization that wants to adopt the cloud en mass, the idea of every development team coming up with their own architecture and spending all of that time and duplication and integration of work is a distraction from ultimately their core mission, which is driving more capability and differentiation in the application for their end customer. So, to be blunt, we actually think the idea of having 15 different databases, ‘the right tool for the job’ is the wrong approach. We think that there’s certain key technologies that most organizations will use for 70, 80% of use cases, and then use the niche technologies for where they really need specialized solutions for particular needs.



Corey: So, if you’re starting off with a general-purpose database then, what is the divergence point at which point—like in my case, eventually I have to admit that using TXT records in Route 53 as a database starts to fall down for certain use cases. Not many, mind you, but one or two here and there. At what point when you’re sticking with a general-purpose database does migrating to something else—what’s the tipping point there?



Sahir: Yeah, I think what we see is if you have a general-purpose database that hits the majority of your needs, oftentimes, especially with a microservices kind of modern architecture, it’s not necessarily replacing your general-purpose database with a completely different solution, it may be augmenting it. So, you may have a particular need for, I don’t know deep graph capabilities, for example, for a particular traversal use case. Maybe you augment that with a specialized solution for that. But the idea is that there’s a certain set of velocity you can enable an organization by building skill set and consolidation around a technology provider that gives much more repeatability, security, less data duplication, and ultimately focuses your organization in teams on innovation as opposed to plumbing and that’s where the 15 different databases been cobbled together may be interesting, but it’s not really focusing on innovation, it’s focusing more on the technology problems that you solved.



Corey: So, we’re recording this on site in Las Vegas, as re:Invent, thankfully and finally, draws to a close. How was your conference?



Sahir: It’s been fantastic. And to be clear, we are huge fans and partners of AWS. This is one of our most exciting conferences we sponsor. We go big, [laugh] we throw a party, we have a huge presence, we have hundreds of customer meetings. So, although I’m a little ragged, as you can probably tell from my voice from many meetings and conversations and drinks with friends, it’s actually been a really great week.



Corey: It is one of those things where having taken a year off, you forget so much of it, where it’s, “Oh, I can definitely walk between those two hotels,” and then you sort of curse the name of God as you wind up going down that path. It was a relief, honestly, to not see, for example, another managed database service being launched that I can recall in that flurry of announcements, did you catch any?



Sahir: I didn’t catch any new particular database services that at least caught my eye. Granted, I’ve been in meetings most of the time, however, we’re really excited about a lot of the infrastructure innovation. You know, I just happened to have a meeting with the compute teams on the Amazon side and what they’re doing with, you know, Wavelength, and Local Zones, and new hardware, and chips with Graviton, it’s all stuff we’re really excited about. So, it is always interesting to see the innovation coming out of AWS.



Corey: You mentioned that you are a partner with AWS, and I get it, but AWS is also one of those companies whose product strategy is ‘yes.’ And they a couple years ago launched their DocumentDB, in parentheses with MongoDB compatibility, which they say, “Oh, customers were demanding this,” but no, no, they weren’t. I’ve been talking to customers; what they wanted was actual MongoDB. The couple of folks I’m talking to who are using it are using it for one reason and one reason only, and that is replication traffic between AZs on native AWS services is free; everyone else must pay. So, there’s some sub-offering in many respects that is largely MongoDB compatible to a point. Okay, but… how do you wind up, I guess, addressing the idea of continuing to partner with a company that is also heavily advantaging its own first party services, even when those are not the thing that best serves customers.



Sahir: Yeah, I’ve been in technology for a while, and you know, the idea of working with major platform players in the context of being, in our case, a customer, a partner, and a competitor is something we’re more than comfortable with, you know, and any organization at our scale and size is navigating those same dynamics. And I think on the outside, it’s very easy to pay way more attention to the competitive dynamics of oh, you run in AWS but you compete with them, but the reality is, honestly, there’s a lot more collaboration, both on the engineering side but also in the field. Like, we go jointly work with customers, getting them onto our platform, way more often than I think the world sees. And that’s a really positive relationship. And we value that and we’re investing heavily on our side to make sure you know, we’re good partners in that sense.



The nuances of DocumentDB versus the real MongoDB, the reality of the situation is yes, if you want the minimal MongoDB experience for, you know, a narrow percentage of our functionality, you can get that from that technology, but that’s not really what customers want. Customers choose MongoDB for the breadth of capabilities that we have, and in particular, in the last few years, it’s not just the NoSQL query capability of Mongo, we’ve integrated rich aggregation capabilities for analytics, transactional guarantees, a globally distributed architecture that scales horizontally and across regions much further than anything a relational architecture can accomplish. And we’ve integrated other domains of data, so things like full text, search, analytics, mobile synchronization are all baked into our Atlas platform. So, to be honest, when customers compare the two on the merits of the technology, we’re more than happy to be competitors with AWS.



Corey: No, I think that everyone competes with AWS, including its own product teams amongst each other because, you know, that’s how you, I guess, innovate more rapidly. What do I know? I don’t run a hyperscale platform. Thankfully.



If I go and pull up your website, it’s mongodb.com. It is natural for me to assume that you make a database, but then I start reading; after the big text and the logo, it says that you are an application data platform. Tell me more about that.



Sahir: Yeah, and this has been a relatively new area of focus for us over the last couple of years. You know, I think many people know MongoDB as a non-relational modern database. Clearly, that’s our core product. I think in general, we have a lot of capabilities in the database that many customers are unaware of in terms of transactional guarantees and schema management and others, so that’s kind of all within the core database. But over the last few years, we’ve both built and acquired technology, things like Realm, that allows for mobile synchronization; event-driven architectures; APIs to be created on your data Easily; Atlas data lake, which allows for data transformation and analytics to be done using the same API as the core Mongo database; as I mentioned a couple minutes ago, things like search, where we actually allow customers to remove the need for a separate search engine for their application and make it really seamless operationally, and from the developer experience standpoint.



And you know, there’s no real term in the industry for that, so we kind of describe ourselves as an application data platform because really, what we’re trying to do is simplify the data architecture for applications, so you don’t need ten different niche database technologies to be able to build a powerful, modern, scalable application; you can build it in a unified way with an amazing developer experience that allows your teams to focus on differentiation and competitiveness as opposed to plumbing together the data infrastructure.



Corey: So, when I hear platform, I think about a number of different things that may or may not be accurate, but the first thing that I think is, “Oh. There’s code running on this then, as sort of part of an ecosystem.” Effectively is their code running on the data platform that you built today that wasn’t written by people at MongoDB?



Sahir: Yes, but it’s typically the customer’s code as part of their application. So, you know, I’ll give you a couple of simple examples. We provide SDKs to be able to build web and mobile applications. We handle the synchronization of data from the client and front end of an application back to the back end seamlessly through our Realm platform. So, we’re certainly, in that case, operating some of the business logic, or extending beyond sort of just the back end data.



Similarly, a lot of what we focus on is modern event-driven architectures with MongoDB. So, to make it easier to create reactive applications, trigger off of changes in your data, we built functions and triggers natively in the platform. Now, we’re not trying to be a full-on application hosting platform; that’s not our business, our business is a data platform, but we really invest in making sure that platform is open, accessible, provides APIs, and functional capabilities make it very easy to integrate into any application our customers want to build.



Corey: It seems like a lot of different companies now are trying to, for lack of a better term, get some of the love that Snowflake has been getting for, “Oh, their data cloud is great.” But when you take a step back and talk to people about, “So, what do you think about Mongo?” The invariable response you’re going to get every time is, “Oh, you mean the database?” Like, “No, no. The character from the Princess Bride. Yes, the database.” How do you view that?



Sahir: Yeah, it’s easy to look at all the data landscape through a simple lens, but the reality is, there’s many sub markets within the database and data market overall. And for MongoDB we’re, frankly, an operational data company. And we’re not focused on data warehousing, although you can use MongoDB for various analytical capabilities. We’re focused on helping organizations build amazing software, and leveraging data as an enabler for great customer experiences, for digital transformation initiatives, for solving healthcare problems, or [unintelligible 00:12:51] problems in the government, or whatever it might be. We’re not really focused on selling customers’—or platforms of data from—not the customers’ data, but other—allowing people to monetize their data. We’re focused on their applications and developers building those experiences.



Corey: Yeah. So, you’re if you were selling customers’ data, you just rebrand as FacebookDB and be done with it, or MetaDB now—



Sahir: MetaDB?



Corey: Yeah. As far as the general Zeitgeist around Mongo goes, back when I was first hearing about it, in I don’t know, I want to say the first half of the 2010s, the running gag was, “Oh, Mongo. It’s Snapchat for databases,” with the gag being that it lost production data was unsafe for a bunch of things. To be clear, based upon my Route 53 comments, I am not a database expert by any stretch of the imagination. Now, the most common thing in my experience that loses production data is me being allowed near it. But what was the story? What gave rise to that narrative?



Sahir: Yeah, I think that—thank you for bringing that up. I mean, to be clear, you know, if a database doesn’t keep your data safe, consistent, and guaranteed, the rest of the functionality doesn’t matter, and we take that extremely seriously at MongoDB. Now, you know, MongoDB, has been around a long time, and for better or worse—I think there’s, frankly, good things and bad things about this—the database exploded in popularity extremely fast, partially because it was so easy to use for developers and it was also very different than the traditional relational database models. And so I think in many ways, customer’s expectation of where the technology was compared to where we were from a maturity standpoint, combined with running an operating it the same way as a traditional system, which was, frankly, wrong for a distributed database caused, unfortunately, some situations where customers stubbed their toes and, you know, we weren’t able to get to them and help them as easily as we could. Thankfully, you know, none of those issues fundamentally are, like, foundational problems. You know, we’ve matured the product for many, many years, you know, we work with 30,000-plus customers worldwide on mission-critical applications. I just want to make sure that everyone understands that, like, we take any issue that has to do with data loss or data corruption, as sort of the foundational [P zero 00:14:56] problem we always have to solve.



Corey: I tend to form a lot of my opinions based upon very little on what, you know, sorry to say it, execs say and a lot more about what I see. There was a whole buzz going around on Twitter that HSBC was moving a whole bunch of its databases over to Mongo. And everyone was saying, “Oh, they’re going to lose all their data.” But I’ve done work with a fair number of financial services companies, and of all the people I talk to, they’re pretty far on one end of that spectrum of, “How cool are we with losing data?” So, voting with a testimonial and a wallet like that—because let’s be clear, getting financial services companies to reference anything for anyone anywhere is like pulling teeth—that says a lot more than any, I guess, PR talking points could.



Sahir: Yeah, I appreciate you saying that. I mean, we’re very fortunate to have a very broad customer base, everything from the world’s largest gaming companies to the world’s largest established banks, the world’s most fastest growing fintechs, to health care organizations distributing vaccines with technologies built on Mongo. Like, you name it, there’s a use case in any vertical, as mission critical as you can think, built on our technology. So, customers absolutely don’t take our word for granted. [laugh]. They go, you know, get comfortable with a new database technology over a span of years, but we’ve really hit sort of mainstream adoption for the majority of organizations. You mentioned financial services, but it’s really any vertical globally, you know, we can count on our customer list.



Corey: How do you, I guess, for lack of a better term, monetize what it is you do when you’re one of the open-source—and yes, if you’re an open-source zealot who wants to complain about licensing, it’s imperative that you do not email me—but you are available for free—for certain definitions of free; I know, I know—that I can get started with a two o’clock in the morning and start running it myself in my environment. What is the tipping point that causes people to say, “Well, that was a good run. Now, I’m going to pay you folks to run it for me.”



Sahir: Yeah, so there’s two different sides to that, first and foremost, the majority of our engineering investment for our business goes in our core database, and our core database is free. And the way we actually, you know, survive and make money as a business, so we can keep innovating, you know, on top of the billion dollars of investment we’ve put in our technology over the years is, for customers who are self-managing in their own data center, we provide a set of management tools, enterprise security integrations, and others that are commercially licensed to be able to manage MongoDB for mission-critical applications in production, that’s a product called Enterprise Advanced. It’s typically used for large enterprise accounts in their own data centers. The flagship product for the company these days, the fastest growing part of the business is a product we call Atlas—or platform we call Atlas. That’s a cloud data service.



So, you know, you can go onto our website, sign up with our free tier, swipe a credit card, all consumption-based, available in every AWS region, as well as Azure and GCP, has the ability to run databases across AWS, Azure, and GCP, which is quite unique to us. And that, like any cloud data technology, is then used in conjunction with a bunch of other application components in the cloud, and customers pay us for the consumption of that database and how much they use.



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: I want to zero in a little bit on something you just said, where you can have data shared between all three of the primary hyperscalers. That sounds like a story that people like to tell a lot, but you would know far better than I: how common is that use case?



Sahir: It’s definitely one from a strategic standpoint, especially in large enterprises, that’s really important. Now, to your point, the actual usage of cross-cloud databases is still very early, but the fact that customers know that we can go, in three minutes, spin up a database cluster that allows them to either migrate, or span data across multiple regions from multiple providers for high availability, or extend their data to another cloud for analytics purposes or whatnot, is something that it almost is like science fiction to them, but it’s crucial as a capability I know they will need in the future.



Now, to our surprise, we’ve seen more real production adoption of it probably sooner than we would have expected, and there’s kind of three key use cases that come into play. One—you know, for example, I was with a challenger bank from Latin America yesterday; they need high availability in Latin America. In the countries they’re in, no single infrastructure cloud provider has multiple regions. They need to span across multiple regions. They mix and match cloud providers, in their case AWS being their primary, and they have a secondary cloud provider, in their case GCP, for high availability.



But it’s also regulatorily-driven because the banking SEC regulations in that country state that they need to be able to show portability because they don’t want concentration risk of their banking sector to be on a single cloud provider or single cloud provider’s region. So, we see that in multiple countries happening right now. That’s one use case.



The other tends to be geographic reach. So, we work with a very large international gaming company, majority of their use cases happen to be run out of the US. They happen to have a spike of customers using their game [unintelligible 00:19:58] gamers using it in Taiwan; their cloud provider of choice didn’t have a region in Taiwan, but they were able to seamlessly extend a replica into a different cloud to serve low-latency performance in that country. That’s the second.



And then the third, which is a little bit more emerging is kind of the analytic-style use case where you may have your operational data running in a particular cloud provider, but you want to leverage the best of every cloud provider’s, newest, fanciest services on top of your data. So, isn’t it great if you can just hit a couple clicks, we’ll extend your data and keep it in sync in near real time, and allow you to plumb into some new service from another cloud provider.



Corey: In an ideal world with all things being equal, this is a wonderful vision. There’s been a lot of noise made—a fair bit of it by me, let’s be fair—around the data egress pricing for—it’s easy to beat up on AWS because they are the largest cloud provider and it’s not particularly close, but they all do it. Does that serve as a brake on that particular pattern?



Sahir: Thankfully, for a database like ours and various mechanisms we use, it’s not a barrier to entry. It’s certainly a cost component to enabling this capability, for sure. We absolutely would love to see the industry be more open and use less of egress fees as a way to wall people into a particular cloud providers. We certainly have that belief, and would push that notion and continually do in the industry. But it hasn’t been a barrier to adoption because it’s not the major cost component of operating a multi-cloud database.



Corey: Well, [then you start 00:21:27] doing this whole circular replication thing, at which point, wow. It just goes round and round and round and lives on the network all the time. I’m told that’s what a storage area network is because I’m about as good at storage as I am at databases. As you look at Atlas, since you are in all of the major hyperscalers, is the experience different in any way, depending upon which provider you’re running in?



Sahir: By and large, it’s pretty consistent. However, what we are not doing is building to the lowest common denominator. If there’s a service integration that our customers on AWS want, and that service doesn’t integrate, it doesn’t exist on another cloud provider, or vice versa, we’re not going to stop ourselves from building a great customer experience and integration point. And the same thing goes for infrastructure; if there’s some infrastructure innovation that delivers price, performance, great value for our customers and it’s only on a single cloud, we’re not going to stop ourselves from delivering that value to customers. So, there’s a line there, you know, we want to provide a great experience, portability across the cloud providers, consistency where it makes sense, but we are not going to water down our experience on a particular cloud provider if customers are asking for some native capabilities.



Corey: It always feels like a strange challenge historically to wind up—at least in large, regulated environments—getting a new vendor in. Originally an end run around this was using the AWS Marketplace or whatever marketplace you were using at any given cloud provider. Then procurement caught on and in some cases banned in the Marketplace outright and now, the Marketplace is sort of reformed, in some ways, to being a tool for procurement to use. Have you seen significant uptake of your offering through the various cloud marketplaces?



Sahir: We do work with all the cloud marketplaces. In fact, we just made an announcement with AWS that we’re going to be implementing the pay-as-you-go marketplace model for self-service as well on AWS. So, it is definitely a driver for our business. It tends to be used most heavily when we’re selling with the, you know, sales teams from the cloud providers, and customers want to benefit from a single bill, benefit from, you know, drawing down on their large commitments that they might have with any given cloud providers. So, it drives really good alignment between the customer, us as a third-party on AWS or Azure GCP, and the infrastructure cloud provider. And so we’re all aligned on a motion. So, in that sense, it’s definitely been helpful, but it’s largely been a procurement and fulfillment sort of value proposition to drive that alignment, I’d say, by and large today.



Corey: I don’t know if you’re able to answer this without revealing anything confidential, so please feel free not to, but as you look across the total landscape—since I would say that you have a fairly reasonable snapshot of the industry as a whole—am I right when I say that AWS is the behemoth in the space, or is it a closer horse race than most people would believe, based upon your perspective?



Sahir: I think in general, for sure AWS is the market share leader. It would be crazy to say anything otherwise. They innovated this model, you know, the amount of innovation happening at AWS is incredible, you know, and we’re benefiting from it as a customer as well. However, we do believe it’s a multi-cloud future. I mean, look at the growth of Azure. You know, we’re seeing Google show up in large enterprises across the globe as well.



And even beyond the three American clouds, you know, we work heavily with Alibaba and Tencent in mainland China, which is a completely different market than Western world. So, I do think the trend over time will be a more heterogeneous, more multi-cloud world—which I’m biased; that does favor MongoDB, but that’s the trend we’re seeing—but that doesn’t mean that AWS won’t continue to still be a leader and a very strong player in that market.



Corey: I want to talk a little bit about Jepsen. And for those who are unaware, jepsen.io is run by Kyle Kingsbury. Kyle is wonderful, and he’s also nuts. If you followed him back when he was on Twitter, you’ve also certainly seen them.



But beyond that, he is the de facto resource I go to when it comes to consistency testing and stress testing of databases. I’m a little annoyed he hasn’t taken on Route 53 yet, but hope does spring eternal. He’s evaluated Mongo a number of times, and his conclusions, as always are mixed sometimes, shall we say, incendiary, but they always seem relatively fair. What is your experience been, working with him? And do you share my opinion of him as being a neutral and fair arbiter of these things?



Sahir: I do. I think he’s got real expertise and credibility in beating up distributed database systems and finding the edges of where they don’t live up to what we all hope they do, right? Whether it’s us or anyone else, just to be clear. And so anytime Kyle finds some flaw in MongoDB, we take it seriously, we add it to our test suite, [laugh] we remediate, and I think we have a pretty good history of that. And in fact, we’ve actually worked with Kyle to welcome him beating up our database on multiple occasions, too, so it’s not an adversarial relationship at all.



Corey: I have to ask, since you are a more modern generation of database, then many from the previous century, but there’s always been a significant, shall we say… concern, when I wind up looking at it [it again in 00:26:33] any given database, and I look in the terms and conditions and, like, “Oh, it’s a great database. We’re by far the best. Whatever you do, do not publish benchmarks.” What’s going on with that?



Sahir: I think benchmarks can be spun in any direction you want, by any vendor. And it’s not just database technology. I’ve been in IT for a while, and you know, that applies to any technology. So, we absolutely do not shy away from our performance or benchmark or comparisons to any technology. We just think that, you know, vendors benchmarking technologies for their—are doing so largely to only make their own technologies look good versus competition.



Corey: I tend to be somewhat skeptical of the various benchmark stuff. I remember repeatedly oh, I’ll wind up running whatever it is—I think it’s Geek Speed—on my various devices to see oh, how snappy and performant is it going to be? But then I’m sitting there opening Microsoft Word and watching the beach ball spin, and spin, and spin, and it turns out, don’t care about benchmarks in a real-world use case in many scenarios.



Sahir: Yeah, it’s kind of a good analogy, right? I mean, performance of an application, sure, the database at the heart of it is a crucial component, but there’s many more aspects of it that have to do with the overall real world performance than just some raw benchmark results for any database, right? It’s the way you model your data, the way the rest of the architecture of the application interacts and hangs together with the database, many, many layers of complexity. So, I don’t always think those benchmarks are indicative of how real world performance will look, but at the same time, I’m very confident in MongoDB’s performance comparatively to our peers, so it’s not something we’re afraid of.



Corey: As you take a look at where you’ve been and where you are now, what’s next? Where are you going? Because I have a hard time believing that, “Yep, we’re deciding it’s feature complete and we’re just going to sell this until the end of time exactly as is, we’re laying off our entire engineering team and we’re going to be doing support from our yacht, parked comfortably in international waters.” That’s a slightly different company. What’s the plan?



Sahir: So, [laugh] you’re—we are not parking anything, anytime soon. We are continuing to invest heavily in the innovation of the technology, and really, it’s two reasons: you know, one, we’re seeing an acceleration of adoption of MongoDB, either with any customers that have used us for a long time, but for more important and more use cases, but also just broader adoption globally as more and more developers learn to code, they’re choosing Mongo as the place to start, increasingly. And so that’s really exciting for us, and we need to keep up with those customer demands and that roadmap of asks that they have.



And at the same time, customer requirements are increasing as more and more organizations are software-first organizations, the requirements of what they demand from us continually increase, which requires continual innovation in our architecture and our functionality to keep up with those and stay ahead of those customer requirements. So, what you’ll see from us is, one, making sure we can build the best modern database we can. That’s the core of what we do; everything we do now especially is cloud first, so working closely with our cloud partners on that. And even though we’re very fortunate to be a high-performance, high-growth company with a very pervasive open technology, we’re still in a giant market that has a lot of legacy technologies powering old applications. So, [laugh] you know, we have a long, long runway to become a long-standing major player in this market.



And then we’re going to continue this vision of an application data platform, which is really just about simplifying the capabilities and data architecture for organizations and developers so they can focus on building their application and less on the plumbing.



Corey: I want to thank you so much for taking the time to speak with me today. If people want to learn more, where can they go?



Sahir: Clearly, you can go to mongodb.com. You can also reach out to us on our community sites: our own or on any of the public sites that you would typically find developers hanging out. We always have folks from our teams or our champions program of advocates worldwide helping out our customers and users. And I just want to thank you, Corey, for having me. I’ve followed you online for a while; it’s great to finally be able to meet in person.



Corey: Uh-oh. It’s disturbing having realized some of the things I’ve said on Twitter and realizing I’m now within range to get punched in the face. But, you know, we take what we can get. Thank you so much for taking the time to speak with me. I appreciate it.



Sahir: My pleasure.



Corey: Sahir Azam, Chief Product Officer at MongoDB. 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 telling me that is not the reason that AWS is building many new databases. Tell me which one you’re building and why it solves a problem other than getting you the promotion you probably don’t deserve.



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.



View Full TranscriptHide Full Transcript