MongoDB Fleshes Out Mobile and Cloud Strategies
As its name implies, MongoDB likes to do big things. But with today’s unveiling of MongoDB Mobile at MongoDB World in New York City, the company is delivering a diminutive database that runs on mobile devices. It also announced the general availability of Stitch, its serverless cloud database environment, as well as version 4.0 of its eponymous NoSQL document store.
As the breadth of depth of data that organizations want to capture and store continues to increase, it exerts ever greater pressure on IT infrastructure. That’s just as true in the data center, where big databases continue to rack up the terabytes and grow outward in a distributed manner, as it is on the edge of the network, where the Internet of Things (IoT) is creating new data-driven use cases by the day.
While the MongoDB database was originally created to run on big clusters of X86 servers, there’s nothing stopping it from running elsewhere. Indeed, customers have been able to run MongoDB on everything from laptops to big IBM mainframes and Power Systems servers for a while. Now, with the launch of MongoDB Mobile, the company is taking the database in an entirely new direction.
MongoDB is filling a need in the market with MongoDB Mobile, according to Seong Park, the vice president of strategy and product marketing for MongoDB. “Mobile devices are something that historically have not been covered very well from a modern database perspective,” Park tells Datanami.
While existing solutions like SQL Lite could be used on mobile devices, Park says, they didn’t give developers the same experience that they have grown to expect with solutions like MongoDB running on cloud or on-prem servers.
Park says the feature set of the new MongoDB Mobile database, which is currently in beta, won’t differ much from its server-based big brother, except that it runs on iOS and Android phones and tablets, as well as other IoT devices. “It will [deliver] the full feature that developers have grown to love and expect in MongoDB into your mobile device, as well as IoT device,” he says.
One of MongoDB’s customers operates oil rigs in middle of ocean, where non-satellite Internet connections do not exist. For them, the capability to capture sensor data and temporarily store it in a mobile database like MongoDB Mobile, and then synch it up to the main server-based MongoDB database when they have a proper connection is a big deal, Park says. “Especially when it comes to equipment failure, they want to be able to get ahead of it,” he says.
Stitch Goes GA
The company’s cloud business has been growing quickly since the company went GA with Atlas two years ago. It has experienced 400% year-over-year growth in cloud revenue, and on the last earnings call, the publicly traded company (NASDAQ Symbol: MDB) said 14% of its total revenues come from Atlas. Out of 6,600 total MongoDB customers, about 4,400 of them are using Atlas.
The company is betting Stitch will send those numbers even higher. Stitch is a serverless cloud database offered under Atlas that the company unveiled as a beta at the MongoDB World in Chicago, Illinois last year. Today the company announced that Stitch is now GA.
Stitch brings four main features: Stitch Trigger, Stitch Query Anywhere, Stitch Functions, and Mobile Synch.
One of the most challenging aspects of managing a mobile database is synchronizing it with the main server-based database when a network connection becomes available. The Mobile Synch function, which will be available soon, will eliminate this complexity, the company says.
The new Stitch Triggers function will allow MongoDB databases to automatically respond to events that it detects by calling Stitch function, which could be utilizing a custom API or a Web hook, Park says. The company will offer native integration with best-of-breed cloud services from AWS or services offered by Twilio or Mailgun, he says.
The new Query Anywhere service in Stitch will essentially allow developers to push queries written in MongoDB Query Language (MQL) out to Web browsers front-ends or mobile devices, without having to worry about security, such as authentication, authorization, and data permissions.
For example, one Stitch customer could use Query Anywhere to simplify how coding around access to medical records, without having to specifically code authentication paths. That could allow a doctor to have full read and write access to a prescription, while restricting patients to read-access, but not giving them permission to write their own scripts or modify exixting prescriptions.
It’s all about taking MongoDB’s guiding principle of “making developers’ lives easier” to the next level, Park says.
“The whole premise of Stitch is to be able to let developers build faster with less code and also provide a fair amount of reusability so that they can use the power of the platform to just build applications faster and obviously have full-fledged feature of MongoDB, as well as best of breed service integration,” Park says.
Today’s MongoDB World conference is also the setting for the debut of another feature called the global cluster.
According to Park, the global cluster makes it easier for customers to configure and operate large clusters spanning multiple regions within a given cloud provider’s environment (the company doesn’t yet offer the capability to run clusters that span cloud platforms, he says).
“That’s been something that’s very much needed,” he says. “If you think about the focus on data governance, data locality, data placement, what global clusters allow you to do is have very low latency read and write capability anywhere in the world, which was not an easy engineering feat to do, especially for a globally distributed database.”
MongoDB supports Atlas on every major cloud platform, including Amazon Web Services, Google Cloud Platform, and Microsoft Azure. The company’s goal is to let customers go to whichever cloud platform they want.
“We want to provide that cloud-provider agnostic experience and unify it through that single Atlas user interface, so we abstract out any of the real difference between cloud providers,” Park says.
On the security front, the company announced the availability of Atlas Enterprise, which will bring several security features, such as support for LDAP integration and support for the AWS Key Management Store (KMS). That will give customers the capability to use the AWS KMS to manage encryption keys that protect data stored in MongoDB.
MongoDB also announced that it’s extending its Atlas Free Tier to the Google Cloud Platform. Customers utilizing this service will get 512 MB of storage, the company says.
It also announced the beta for a MongoDB Kubernetes Operator, which supports the provisioning of stateful, distributed database clusters and coordinates the orchestration between Kubernetes and MongoDB Ops Manager.
This feature, which MongoDB first unveiled in February, lets customers build transactional applications that feature all the atomicity, consistency, isolation, and durability, or ACID, properties that are needed. The company’s storage engine previously supported ACID transactions in single-document transactions, but not with transactions that touch multiple documents.
While there is a performance hit when using the multi-document ACID feature of MongoDB’s WiredTiger database engine compared to not using it, there’s a good chance that its faster than the multi-document transaction processes that customers built themselves. According to Park, one such user found MongoDB 4.0 was 2x faster than their own hand-written code.
Lastly, the company also announced that MongoDB Charts, a native graphing functionality in the database. MongoDB says Charts is “the only visualization tool designed to natively handle MongoDB’s rich data structures,” which ensures that Charts gives “the most up-to-date view of live data.”
Charts will be included with MongoDB 4.0, which is now generally available.