Interested in speaking at MongoDB World 2022? Click here to become a speaker.
HomeLearnMichael Lynn

Michael Lynn

Michael Lynn is Principal Advocate at MongoDB. Previously, Michael worked as a Solutions Architect at MongoDB, helping users optimize MongoDB for scale and performance. Earlier, Michael was Vice President and Global Head of UNIX Engineering at a Fortune 50 Financial Services firm where he managed application and infrastructure platform engineering for the UNIX distributed server environment. Prior to that, Michael worked as an application developer, database administrator and as an infrastructure engineer across various companies and organizations.

Articles by Michael Lynn

article

Create a Data Enabled API in MongoDB Realm in 10 Minutes or Less

Learn how to create a data api with MongoDB Realm in 10 minutes or less

podcast

Ep. 91 Loan Laux - MongoDB and E-Commerce

With Black Friday upon us, let's take this opportunity to focus on solutions in the retail and E-Commerce space. Loan Laux is a freelance developer, and the founder of out:grow, an E-Commerce Agency specializing in helping retailers grow, and migrate to more efficient technology solutions. Loan shares the story of his journey as a developer, as a founder and why MongoDB is an essential and efficient part of the modern E-Commerce technology stack.  To contact Loan, send him an email: llaux@trellis.co You will reach Loan on Twitter at https://twitter.com/loanlaux Be sure to stop by the MongoDB Booth at AWS re:Invent. 

podcast

Ep. 90 John Page and Karen Hualme of MongoDB Developer Relations

MongoDB .Local is a great conference concept featuring uniquely hybrid offerings including education, keynotes and product updates, and entertainment curated for those joining live in person, live from home, or on-demand. The latest .Local was held on November 9th, 2021 at Evolution, London. In this episode, Karen Hualme joins the growing ranks of MongoDB Podcast guests and shares her experience as a Principal Developer Advocate. Next, John Page stops by to talk about his favorite announcements from the conference and to share details of the amazing projects he's working on for the MongoDB Community. This episode was brought to you, in part by the MongoDB Engineering team. We're hiring! If you've thought about joining the team that's working hard to improve on the best data platform on the planet, visit https://bit.ly/mongodbengineering to find out more information and apply.  MongoDB Conferences are a great way to connect with the community, learn from others doing what you're doing and they provide a great opportunity to share what you know. MongoDB World Call for Speakers is open! If you want to learn more about speaking at a MongoDB Conference, visit https://bit.ly/speakatworld. If you're concerned that you might not be ready, be sure to check out Episode 75 of this podcast to learn all about the MongoDB Speaker Program.  

podcast

Ep. 89 Richard Collins and DJ Walker Morgan of Studio3T

MongoDB .Local is a great conference concept featuring uniquely hybrid offerings including education, keynotes and product updates, and entertainment curated for those joining live in person, live from home, or on-demand. The latest .Local was held on November 9th, 2021 at Evolution, London. In this episode we chat with Richard Collins, CMO and DJ Walker-Morgan of 3T Software - makers of Studio3T. Studio 3T is the professional IDE, client, and GUI for MongoDB. The right tools to get more done in MongoDB – on Atlas or anywhere. To try Studio 3T for free, visit https://studio3t.com/download/  

podcast

Ep. 88 Harry Swan and Cademy

Cademy is a hyper-local educator marketplace that enables content and course creators to more easily administer and market their courses. Reducing the time and effort spent on admin so they can focus on what truly matters - teaching and making your students happy!  Harry Swan is the CTO, and Co-founder of Cademy. He stopped by the podcast booth at MongoDB .Local London to share details of his journey with MongoDB and his thoughts on some of the latest innovations presented at the conference. 

podcast

Ep. 87 Matt Asay and Joe Drumgoole Live at .Local London

MongoDB .Local is a great conference concept featuring uniquely hybrid offerings including education, keynotes and product updates, and entertainment curated for those joining live in person, live from home, or on-demand. The latest .Local was held on November 9th, 2021 at Evolution, London. In this, the first in a series episodes recorded live at the event, we chat with two industry veterans, Matt Asay, and Joe Drumgoole.  Matt and Joe share their experience as long time (and in Matt's case, returning) MongoDB employees as well as their thoughts on the latest announcements from the event. 

podcast

Ep. 85 Tech Conferences with Nancy Monaghan and Dorothy McClelland

Here we are in November of 2021. The pandemic continues to impact the way we’re all working, doing business and connecting with other people. One powerful way that technologists from all industries have connected is through attendance at conferences. I’m no different… I find attending a good tech conference to be a valuable part of my continuing education. From large scale conferences like the Consumer Electronics Show, and AWS Re:invent, to language-aligned conferences like Gophercon, and PyCon. I go because it’s an opportunity to meet professionals and gain valuable perspective… and learn about the latest updates from vendors and technology providers.  On today's episode, we’re focusing on Tech Events and Conferences and to help guide a discussion about this and about the changing landscape of tech events throughout the pandemic, I’ve invited two of my highly esteemed colleagues from MongoDB’s amazing Events Management team… Nancy Monaghan and Dorothy McClelland.  If you’re a technologist interested in the new and exciting ways, we at MongoDB are approaching hybrid conferences or if you’re considering changes to the way you run and host events, this episode is for you. 

podcast

Ep. 84 Shrey Batra - Writing Scalable Applications for Big Data with MongoDB

Shrey Batra is an engineer at LinkedIn. Shrey is a long-time member of the MongoDB Community and has spoken at MongoDB Conferences and led user group activities for quite some time.  Shrey joins Michael Lynn today to talk about writing scalable applications for Big Data. Shrey stresses the importance of data modeling, and shares techniques that he has leveraged to ensure that applications can not only manage large volumes of data but that can scale over time as use patterns grow and change.  Resources: Data Model Examples and Patterns: https://docs.mongodb.com/manual/applications/data-models/ MongoDB University - M320 Data Modeling: https://university.mongodb.com/courses/M320/about  

podcast

Ep. 83 Online Archive and Data Lake with Benjamin Flast

With MongoDB Atlas Online Archive you can automatically tier your data across fully managed databases and cloud object storage while preserving the ability to query all your data through a single endpoint. Create a rule to automatically archive infrequently accessed data from your MongoDB Atlas database to optimize cost and performance. On today's episode, we chat with Benjamin Flast, Sr. Product Manager for MongoDB Atlas - Online Archive and Data Lake.

podcast

Ep. 82 Serverless with Chris Shum

Serverless computing, pioneered by platforms such as AWS Lambda, is a relatively new computing paradigm that developers are increasingly using as a tool for building applications in the cloud. Generally speaking, a solution is serverless if it can automatically and dynamically match compute and storage resources to workload demands. On today's show, Chris Shum, Sr. Product Manager for MongoDB Serverless (available in preview) joins Michael to talk about what Serverless means in the MongoDB context, how it was created and why you should be interested. # Chris Shum - Serverless [00:00:00] **Chris Shum:** Hi everybody. My name is Chris Shum and I work on MongoDB Atlas, the most advanced cloud database surface of the market. Stay tuned to learn all about how MongoDB is thinking about scalability for the next generation of modern applications and how we're building a new serverless offering that might save you hundreds of hours. [00:00:20] **Michael Lynn:** Welcome to the show on today's episode, Christian. Senior product manager, working on Mongo, DB, Atlas, serverless instances. It's a brand new offering. What is serverless? You might be asking, you're familiar with the term probably using it today, but what does it mean in the context of MongoDB? Stay tuned to find out. I hope you enjoy this episode. Chris, welcome to the podcast. It's great to have you on the show. How are you doing today? [00:00:49] **Chris Shum:** Thanks for having me on like I'm doing great. [00:00:51] **Michael Lynn:** Fantastic. Great. So we're going to talk about a lot of things, but I want to get right into the serverless topic. Serverless is something that's confusing in many ways, there are servers, right? So it's not truly serverless, but I think it would be great if we started out the episode with maybe your take or your definition of what serverless is. [00:01:13] **Chris Shum:** Yeah, absolutely. And I don't blame you at all. I think many of us are all scratching our heads asking the same question. The way that I think about it is really that serverless is a relatively new computing paradigm. The developers are increasingly thinking about and using as a tool for building applications in the cloud. And generally speaking, I think a solution would be deemed serverless if it can automatically and dynamically scale compute and storage resources to meet workload demands, it's a pretty broad definition, but I think. Ultimately, it's an abstraction that frees developers from needing to think about servers and serverless. So you're absolutely right. It's not that surface don't exist. It's just the developers don't have to think about. Ah, so it [00:02:02] **Michael Lynn:** could more aptly be called headache list [00:02:05] **Chris Shum:** perhaps. Yeah. Maybe we should probably do some marketing and try to see if that right. [00:02:12] **Michael Lynn:** Tell me a little bit about yourself, Chris. And how long have you been at MongoDB? [00:02:17] **Chris Shum:** I am a senior product manager at MongoDB. I've been working on MongoDB Atlas for over four years now. And it's been an incredibly exciting time to just see more and more folks be able to use Maga DB, enjoy the benefits of the flexible data, flexible document model and really such a joy to see Atlas the platform take off. I recently had a hand in helping launch Atlas serverless instances this past summer@mongodb.live 2021. So I am incredibly excited and honored for Mike to feature. We are happy on his podcast to be able to talk a little bit more about. [00:02:53] **Michael Lynn:** Fantastic. And it's truly great to have you on and get to the bottom of some of these interesting developments that we've got going on. And so for years, that's essentially from the beginning of Atlas, right? You were here right around the time that Atlas law. Yeah, [00:03:07] **Chris Shum:** Atlas just celebrated its five-year anniversary. So I missed kind of the first year where we quote unquote Atlas learned how to walk, but I've seen it grow really from a product that no one knew about. And in the earlier days, our struggle was really trying to get the brand of Margaret to be Atlas out there. And today we have many customers who'd never even thought or knew about Margo. Not having a cloud offering. So it's incredible to see the advancements that we've made in only just five years. And today MongoDB is 56% of MongoDB, the company's revenue where have over 26,000 customers running on the platform over 2 million databases. So we're incredibly excited. And I think this is honestly only just the start. Of what we're capable of and what we want to do with going to be Atlas it's been a great pleasure. [00:04:01] **Michael Lynn:** I do want to find out a little bit more about your journey and how you ended up at MongoDB. What did you do prior to coming to MongoDB? [00:04:08] **Chris Shum:** I was in school prior to joining MongoDB. I actually interned here as a summer intern my junior year, and I never looked back so to speak. I. I went to brown university in Providence, Rhode Island. We actually have a very large brown faction here at MongoDB. Our, one of our co-founders Elliot Horowitz was was at brown university. And I heard many great things from from many of the brown alumni who work at MongoDB. And so I decided to just join for a summer and see what it was like. And it's been the ride of a lifetime. I can't say I've not that I've worked many other places. I certainly have interned at smaller companies, but my DB just has such a. An incredible way of doing product development. It's such a great culture. It's got, and I enjoy the people that I work with every day where everyone's incredibly humble. While we work on some of the hardest problems that people in the industry are trying to solve. [00:05:03] **Michael Lynn:** What did you study in school? What led you to this [00:05:06] **Chris Shum:** path? Yeah at brown, I did a double major in computer science and economics. And so product management was probably the right fit for me. I after a while, but I actually got my start at market DB as a software engineer, a building Mongo, DB Atlas as an engineer that was incredibly fun solving, just some incredibly complex technical challenges. But after I think it was about 18 months, maybe 20 as an engineer, I realized that I was never going to be the best engineer. And I probably had a little bit more curiosity interests outside of engineering that I wanted to try and satisfy. And that's how I got my start in product management. And so since since transitioning from being a software engineer have had the opportunity to work on still the same domain and the same problem set. That is how do we get Margaret to be Atlas, to continue becoming the most resilient, the most flexible cloud database data platform, but at a slightly different angle, right? Thinking about what sorts of. No go to market challenges, commercialization, packaging, listening to our customers, incredibly attentively, to understand what are the pains that they're, what are the pains they're having and making sure that we're building a product that is attuned to the needs and pains that they have, and making sure that we're solving their problems. It's taken probably a very large village to build out lists and to build it to what it is today. And I certainly cannot claim any credit for the success of what Atlas is today, but it's. I'm incredibly fun to just observe how this rocket ship is taking off [00:06:45] **Michael Lynn:** now. So I'm going to be the folks that are listening that are familiar with going to be as a, just a straight database. It started its life. As an open source project, you can download it, install it on your laptop, develop for free launch your apps. Run them for free on your laptop on your server. You can install it in a data center. And it was largely successful even before the concept of cloud came to be, I'm wondering if you think that the success of Atlas is largely due to the success of Mongo to be as a database, or do you feel like the market trends really pulled Mongo to be in that direction? [00:07:21] **Chris Shum:** That's a very good question, Mike. I think. I think MongoDB owes a lot of its success to the open source community and the very large community of developers that rallied behind what was frankly, a very disruptive database model in the market that had been stagnant for many years. It was a model that few people thought was. What succeed and w honestly it could be used for such kind of general purpose, mission, critical workloads. And I wouldn't say that Margaret DB success is owed to the cloud migration or that MongoDB was pulled by the cloud migration. I like to think about it as. MongoDB really being true to its vision of making data stunningly, easy to work with and therefore being incredibly attuned to what are the developers at large wanting to do, or where are they going? I think that our. Cloud offering. And even now developing an incredibly robust application data platform is the result of listening to and working with. Tens of thousands of developers out there who are many of them today with they start a project, don't even think about doing an M doing anything on premise or locally, what is even an on-prem data center anymore. Everyone's just starting to do things on the public cloud, leveraging cloud native building blocks. And MongoDB realized that very early on was that cloud was going to be the next big thing that. If we were going to continue serving our developers was incredibly important that we had a compelling solution there. So I think that's honestly that's how we got got into the cloud. And generally it's been fortunate to grow as public cloud adoption has significantly increased. [00:09:21] **Michael Lynn:** So really just a natural extension to the original mission of making data, easy to work with from a database perspective. Now, extending that. To today's paradigm where we're all using cloud services. And it literally seems like everything we do is SAS based. [00:09:39] **Chris Shum:** Absolutely. Absolutely. We're going to talk a little bit [00:09:42] **Michael Lynn:** about serverless and I think it might be helpful if you describe the product what is Mongo to be serverless and why was it. [00:09:52] **Chris Shum:** Yeah, it's a very good question. So I think about it in a couple of dimensions. So maybe the first one we want to tackle is why might developers care about serverless? And so one of the. Problems. And some of the problems and challenges that developers often take about is in terms of infrastructure sizing and database capacity management. And I did a little bit of homework prior to joining this recording session. And I found out that developers there's some studies out there that that say developers spend about 41% of their time on infrastructure bank. And that's a lot of time as a developer who really wants to be working on the next, the next big thing, the next unicorn, that's a lot of time to be spent thinking about maintaining infrastructure that we think about as non-differentiating work. And even though that we spend a lot of time maintaining infrastructure, we don't always get it right. And there are a lot of consequences of incorrectly sizing. Infrastructure right there. There's number one is there's management overhead. Sometimes you just, aren't confident in your sizing and therefore you'll feel the constant need to monitor resource consumption and make sure there's sufficient capacity. If you size it incorrect, even if you've sized it correctly, initially your application might change in workload patterns. You might hit the front page of hacker news and blow up and, you need to go immediately, go back and adjust your resourcing accordingly. And then there's the worst case if you've incorrectly sized and you've under provisioned, your application might all of a sudden become slow or become unresponsive. If you hit some sort of resource bottles. And let's say, I'm going to pay for insurance and I'm going to over-provision and now you're just paying for unused capacity, which can quickly become expensive. And as an engineer at MongoDB, there's always some combination of these three things that kept me up at night. And so those problems and concerns around infrastructure sizing and capacity management. It is a big motivation for why serverless becomes interesting to developers. I think of it as serverless to delivering on two main tenants. Number one is elastic scaling, right? The ability to automatically scale up and down based on workload, including sometimes the ability to scale down to zero resources when there's absolutely no workload. The second tenant is. Consumption-based pricing, right? A pricing model that only charges for resources used to surface the workload. And if I played those tenets back in maybe more user friendly words, that model has two main benefits. Number one, you don't need to think about scaling up to meet increasing workloads, and you don't need to worry about paying for resources. You're not using it's a serverless is a model that dynamically uses only what it needs. And only charges for what it uses. One analogy that was incredibly helpful for me, as I started getting deeper into serverless is power utilities. When we're home, we turn on the air conditioner or flip on the light switch without ever thinking about the amount of natural gas that a power plant is burning to deliver our power. Instead, we expect to get a fair bill at the end of the month, commensurate with whatever electricity usage. I never think about the amount of natural gas that's been burned, that would be crazy. And serverless computing and serverless databases strive for the exact same thing. A simple. And you're confident, then you're getting built for what you're consuming while serverless, initially gained popularity on the application layer. We obviously know by AWS Lambda, Google functions, et cetera, we're seeing this wave quickly spread to database technologies as well. And for very good reasons as we've discussed. And so more and more, what MongoDB has observed is that developers are increasingly expecting the database of their choice should be serverless compatible as well. And so that brings us to what we've just announced@margarettobe.live 2021. Now diving into why did the MongoDB decide to build serverless instances? I think that really dovetails with what we were discussing earlier, which is just what is our mission at MongoDB it's to make a data stunningly, easy to work with. And that manifested when we built Atlas, right? We wanted Atlas to, we wanted Atlas. Enable us to deliver the flexibility of the document data model and the best application data platform into the hands of as many developers as possible. And serverless is just a natural extension of that journey. We started Atlas five years ago, making just MongoDB, the core transactional database available as a managed service on a single cloud provider. And since then over the last five years now, we've expanded it to be available on three of the largest public cloud providers, AWS, Google cloud, and Azure. And we're now available in over 80 regions globally. And we. Been over time, expanding from offering MongoDB, just as serving as the beating heart transactional database behind core mission, critical applications into a platform that services many more additional use cases with our integrated search capabilities, mobile and computing at the edge. Real-time analytics, querying data in different sorts of environments like cold storage and much, much more. And. When we were, when we went to think about what's the next step and our application data platform, we asked ourselves what would make it even easier for developers to use Mongo? DB serverless was obviously the answer, right? We wanted to introduce even more flexibility. We wanted to allow customers to deploy instance with almost no configuration required. And we also wanted to expand our ability to meet new workloads. With serverless instances, we now have a range of deployment options to meet a variety of different workloads, and we can better meet different workloads that have, some of them might have infrequent or sparse traffic that aren't an ideal fit for the kind of more traditional database consumption models. And all of these things combined of just where we're seeing the trend, what we were seeing, developers using just a natural segway of where we wanted my BDB to go in our journey. It just made sense to do serverless instances. [00:16:48] **Michael Lynn:** This episode is brought to you by MongoDB dot local London. It's happening this November. Tuesday at evolution park, Battersea park, London, England. It's a uniquely hybrid experience offering education, exploration, and entertainment. Curated for those joining live in person at evolution London live from home or on demand on your own schedule. To get more information, visit MongoDB.link/london, local. I hope to see you there. We've been heading in this direction for some time. And there have been enhancements to Atlas to increase the flexibility with regard to scale. And we introduced a scalable storage first. So you didn't have to think about increasing the amount of disc you assigned to your instances. You could automatically provision a dif additional disc space. And then that was followed up by. Elastic scalability with CPU and memory. And largely you can set a high watermark and a low watermark with elastic, scalability, and essentially there's a radio button that you can enable to scale between those. How does that differ? How does elastic scalability in MongoDB Atlas differ from serverless [00:18:10] **Chris Shum:** elastic, scalability. Always be incredibly important to Atlas in so far as Atlas dedicated clusters are incredibly important in our portfolio and will never go away, but dedicated clusters, we think it's a different solution to. The same class of problems or pain. And obviously with the, with a different solution, there are just different trade-offs right. With a dedicated cluster that gives us a dedicated virtual private cloud VPC. Where w where we'll deploy your cluster into. And that obviously comes with great levels of isolation and security guarantees. You're guaranteed to be in an isolated environment where there are no kind of noisy neighbors or the risk of any data leaking, et cetera. And you can set your own incredibly fine-grained backup and restore policies. And so with data with dedicated clusters, That's where storage scaling and the elastic CPU compute scaling makes a ton of sense. And we will continue to invest in that and make it even more responsive, continue tweaking our algorithms and what sort of metrics that we're scaling off of. But. There will always be trade-offs in that dedicated clusters and how they scale are always subject essentially to the laws of physics. There's a limit to how fast we can scale them. And in truth, it's a problem that regardless of how we decide to tweak the algorithm or introduce machine learning, et cetera, isn't something we can solve elegantly. And that's where serverless comes into the picture. So serverless introduces a brand new. Different architecture, still trying to solve the same pain, but with a slightly different approach to it. And that means that we're able to scale much more seamlessly. We're able to do sort of consumption-based pricing, but there are going to be trade-offs in the. The types of, for example, granularity back granular backup policies you can apply to a serverless instance because it's just built on a slightly different architecture. [00:20:36] **Michael Lynn:** Okay. So I understand what happens in Atlas. When I use the web UI to launch an instance of Mongo to be out. Launch an instance of MongoDB in Atlas. And I select a tier. I configure the instance in a way I go, and after a couple of minutes, I get the cluster instance back and I can connect my application. Do you want to talk a little bit about what happens in the serverless case? [00:20:59] **Chris Shum:** Absolutely. So when we went out to build serverless instances, one thing that we wanted to make sure it was that deployment was going to be safe. So what you do today, when you want to deploy a serverless instance is you can navigate to the Atlas console. That's cloud dot Mongo, db.com. If you don't have an account, please go register it's free, but you go to your Atlas console and all you do is you hit the create button and where in the past, you could only select from either our shared to your offering our dedicated clusters. Today. There's a new option. That is a serverless. And you'll see that if you click into serverless instance, the options are incredibly limited. All you do today is you pick a cloud provider and you pick a region and you give it a name and that's it. There are only three choices that you have and you deploy. It's done incredibly quick to deploy. You as a customer, you no longer have to think about what cluster tier or storage you need for your database, both at deployment time, as well as then on an ongoing basis. You just simply create kind of a cloud database. Think of it as an endpoint in the cloud, and it's going to automatically and dynamically scale to meet your workload demand. And it's a great thing about it is that. Even though it's so simple, it continues to be a fully managed database. It's built on top of the same foundations as an Atlas dedicated cluster. And therefore it comes with many of the same built in best in class operations and features of a fully managed database, including. And to insecurity continuous uptime and backups, automatic upgrades, IX metrics, and alerts, index suggestions. So many more things we've built into Atlas over the years. And since it's built on the same reliable foundation, you don't have to think about management of your serverless instance, whether it's in workload demands or even just reliability of those resources. And now dovetailing back into the two main tenants of serverless that we were describing. And all of this comes with just an incredibly intuitive operations based pricing model. It's a pricing model that we've never done in the history of Mako DB. You only pay for what you use and in the case and that, and we define what you use in incredibly developer friendly terms in that we charge based on reads and writes and storage, right? These are logical units that you're thinking about as you develop your application. And so hopefully that means gone are the days where you're trying to rationalize, Hey, what is 1% of CPU equate to in the number of operations that I can send out to a database instead it's incredibly intuitive. We'll charge you based on reads and writes. And we were trying to innovate on that lens to make it something that's incredibly fair and intuitive at the same. [00:24:07] **Michael Lynn:** Great. It sounds like the ideal situation, but as you're describing it, I'm starting to think about the ideal workloads. So if I'm a developer and I'm ready to launch my app what should I look to in my app to determine whether or not my use cases ideal for serverless [00:24:24] **Chris Shum:** Atlas, surplus instances, number one, they're a brand new serverless database offering in the market. And number two, they're available to the public in a preview capacity or in a public preview capacity. So what that means is MongoDB continues to still be building on top of serverless instances and improving them. And that we're also eagerly learning from our customers. And what that kind of means in terms of our recommendations is that. Today, we'd love to see you put more kind of infrequent or sparse workloads that fit kind of the serverless mental model and do development and sandbox testing against your serverless instance. We're likely still a little bit too early from you putting a mission critical production workload on a serverless. Not that I personally don't have confidence in what we're offering as a serverless instance, but more so because it's still very young and it hasn't been battle tested yet. And I'm sure my. My illegal counterparts, if they were listening to this podcast we are also going to remind me to tell you that serverless instances, because they're still under pre-field capacity. They're not included in our uptime SLA, and therefore they're limited or restricted in functionality. However, this is something that we are incredibly excited about continuing to. I personally would love to hear from you as an audience, if you have any thoughts or feedback, or if you're an early adopter of serverless instances, please feel free to shoot me a message I'd love to hear from you. [00:26:03] **Michael Lynn:** Yeah. And that's great. I would love for the audience to get involved as well. And as Chris mentioned, if you want to try Atlas and serverless, you can jump on over to cloud dot Mongo, db.com. And register for free. It is a free service. You can use MongoDB Atlas. Even in production for free, you don't have to pull out a credit card. The free instances are limited in the capacity. So obviously you're gonna want to take a look at the. The amount of Ram and CPU available to those for your instances. And when you're ready, you can pull out a credit card and increase the capacity associated with your instances. So I'm compelled to ask how then would folks that have given serverless a try, how would they give you some feedback? How do they get in touch? So [00:26:50] **Chris Shum:** number one, I'm actually personally reaching out to every single customer who has been using surplus instances in this early previous stage. And so hopefully if I'm doing, if I'm doing my job right, then you will receive an email in your inbox. Otherwise we do have both a MongoDB feedback engine, which is feedback dot Mongo, db.com, as well as it Margaret, to be community forum, which has a growing group of Margaret to be enthusiastic. So please I personally frequent both of those places. And so if you have comments or feedback, please do leverage those to other forums. Great. [00:27:31] **Michael Lynn:** Yeah, I hope we get some folks jumping in and giving some feedback. I think that's how we're really going to ensure that the service continues to grow in the right way in the ways that developers are really concerned about. And with that in mind, do you want to talk a little bit about the future what's in store for. [00:27:47] **Chris Shum:** Yeah, I am. I am incredibly excited for both serverless as a development practice and a technology that we use very often, as well as for the future of Atlas serverless instances. We at MongoDB believe that the serverless model will become more and more popular as a modern application trend. And that serverless databases are eventually going to be as attractive and as valuable as more of the traditional deployment models. And frankly, it's because serverless realizes the problem is. Of the cloud, right? We've been talking about cloud as something that will help us abstract and automate more and more of the low level infrastructure configurations that developers and engineering teams don't want to have to learn and, or to maintain. And sometimes that is the case for cloud, but for those of us who have managed Infrastructure as a service VMs and gotten into do kernel or OSPF patches. We know that sometimes it's still also a little bit difficult. And so serverless we think is really that next step in terms of actually realizing the promise that has been made by cloud all along. And so when I think about the future it's honestly a multi-year investment for us as a company, and we already have a multi-year vision. As I'm recording this podcast, there are many engineering teams who are working day to day on improvements for serverless instances. As we set a generally available release of serverless instances in our near term horizon. And we're going to continue iterating on the product. While MongoDB as a cloud database were born in the cloud era. There's honestly even more that we can do to adopt cloud data, building blocks and cloud technology breakthroughs that have been made in the last couple of years, as we iterate on our own backend architecture it's something that we're incredibly excited about. It's going to be. An incredibly complicated technical problem and challenge to solve it, which gets me super excited. And I'll probably end by saying it's going to be a marathon, not a sprint. It'll take us a while. I think, to fully achieve the vision that we have for serverless instances, but we're already incredibly excited to foot what we've brought to the market. We're excited for you to give it a try. And fruit to, to give us feedback so that we can continue listening and to learn and to build, as we touched on earlier, I think MongoDB both as a company, as well as the suite of products that we've built are built on the feedback of our community and of our users. We're always eagerly listening and we included. All of the things that are the feedback that we heard into that we here into our roadmap and into our plans. [00:30:54] **Michael Lynn:** One of the things that I really love about going to be Atlas came on fairly early, was access to to the features and functions within Mongo, DB, Atlas via the API and the Atlas API is serverless available as an [00:31:07] **Chris Shum:** API call. Yes. We love our developers who enjoy using infrastructure as code using the CLI the Terraform and cloud formation. Number one, generally, I'll say MongoDB Atlas has a declarative API where you can do everything that you can do in the console via the API. We also have our own command line interface, CLI we have our own Terraform and CloudFormation integrations. And so for those folks who don't like to use. Atlas is the console are still waiting for us to build dark mode into the Atlas UI, and certainly use our API and serverless instances are no different. You can create and manage your surpluses since via our declarative API. [00:31:52] **Michael Lynn:** Now, we talked about the ideal scenario and we talked a little bit about the preview mode. Are there cases where you probably would suggest we don't use. [00:32:02] **Chris Shum:** Honestly, we've been shocked by what serverless instances are capable of ever since we've launched them. Every week we continue to see in customers put larger and larger, more and more active workloads onto serverless instances. And so that might be some. And so we've been shocked in terms of. How scalable serverless is and it's is, can be in that we've observed customers be able to go up to tens of thousands of operations per minute, millions of documents scanned per operation a little bit too easily. And so that might be those where I say It's not that serverless instances, aren't capable, but I'd like to get a little bit more time to see surplus incidences bake and be test is as we incrementally improve and broaden the set of use cases that we would confidently serve on serverless. This is before you throw large production workloads on it. [00:33:04] **Michael Lynn:** Yeah. That makes sense. Where are you looking for inspiration? What companies out there are doing interesting things in the serverless space. Compelling and that you might like to see in our offering. Yeah. There, [00:33:15] **Chris Shum:** there are honestly a lot of companies that are innovating in the serverless industry. If they're of course the cloud providers, and I think each of the cloud providers have their own serverless offerings. There are also incredibly exciting developments, both for Both on kind of the application stack. We know we work closely with Marcel and Netlify is there also developments in the database world as well? For example, five DB and planet scale DB. These are all folks, I think who are doing. Very exciting and interesting stuff in the industry. And we'd love to continue learning from them and improving the adoption of serverless as a technology, across as many developers as possible. Great. [00:34:05] **Michael Lynn:** Yeah. And I love some of those companies. We'll be having an episode shortly where we're talking about. The development side, not necessarily the data, but the development side of serverless. So stay tuned for that. Chris, this has been a great chat. I appreciate you taking the time out of your busy day to, to enlighten us about serverless. Is there anything else you'd like to share with the audience? [00:34:27] **Chris Shum:** Thank you so much for having me, Mike, the only thing that I'd like to echo one more time is. We'd love to see you try out serverless instances. It's available to the public today. It's available on every outlet console or API. So please do go give it a try and share any feedback that you have for us. Thanks very much. Thank you. [00:34:52] **Michael Lynn:** Thanks so much to Chris for joining us today, explaining all the ins and outs of MongoDB serverless instances in Mongo to be Atlas you want to get started head on over to clown dot Mongo, db.com. Sign up. It's free to get started. No credit card required. You can check out serverless. Make sure you click on the feedback link and leave your feedback for Chris. He'll get in touch with. Also remember dot local is coming November 9th. It's going to be in London this time around. If you want to find out more information about dot local London, visit MongoDB.link. That's Mongo, DB dot L I N K slash London. You can get more information there, sign up. And if you are going to be onsite, make sure you stop by the podcast booth. I'll be broadcasting doing some interviews there. We'd love to chat with you about what you're building and your experience with MongoDB. [00:35:42] **Chris Shum:** Have a great day, everybody.

article

At the Intersection of AI/ML and HCI with Douglas Eck of Google (MongoDB Podcast)

Douglas Eck is a Principal Scientist at Google Research and a research director on the Brain Team. His work lies at the intersection of machine learning and human-computer interaction (HCI). Doug created and helps lead Magenta (g.co/magenta), an ongoing research project exploring the role of machine learning in the process of creating art and music. This article is a transcript of the podcast episode where Anaiya Rasinghani leads an interview with Doug to learn more about the intersection between AI, ML, HCI, and Databases.

article

Atlas 5-Year Anniversary Podcast Series Episode 1 - Onramp to Atlas

My name is Michael Lynn, and I’m a developer advocate at MongoDB. I’m excited to welcome you to this, the first in a series of episodes created to celebrate the five year anniversary of the launch of MongoDB Atlas, our Database as a Service Platform. In this series, my co-hosts, Jesse Hall, and Nic Raboy will talk with some of the people responsible for building, and launching the platform that helped to transform MongoDB as a company. beginning with Episode 1, the On ramp to Atlas talking with Sahir Azam, Chief Product Officer, and Andrew Davidson, VP of Product about the strategic shift from a software company to a software as a service business. In episode 2, Zero to Database as a Service, we’ll chat with Cailin Nelson, SVP of Engineering, and Cory Mintz, VP of Engineering - about Atlas as a product and how it was built and launched. In episode 3, we’ll Go Mobile, talking with Alexander Stigsen, Founder of the Realm Mobile Database which has become a part of the Atlas Platform. In episode 4, we’ll wrap the series up with a panel discussion and review some of our valued customer comments about the platform. Thanks so much for tuning in, please take a moment to subscribe for more episodes and if you enjoy what you hear, please don’t forget to provide a comment, and a rating to help us continue to improve.

Load more
MongoDB logo
© 2021 MongoDB, Inc.

About

  • Careers
  • Investor Relations
  • Legal Notices
  • Privacy Notices
  • Security Information
  • Trust Center
© 2021 MongoDB, Inc.