MongoDB SaaS Version, Native Spark Connector Unveiled

9 Tech Giants Embracing The Open Source Revolution

MongoDB is introducing a new elastic database-as-a-service cloud offering for its open source NoSQL database, available initially on Amazon Web Services (AWS), with future integration planned for Google Public Cloud and Microsoft Azure. Separately, the company is also announcing a new MongoDB connector for big data technology Apache Spark.

Called MongoDB Atlas, this on-demand cloud service will provide an outsourced and managed version of the NoSQL database. It's aimed at a broad market, from small players to MongoDB's Fortune 100 customers. Both the MongoDB-as-a-service and Spark connector announcements coincide with the company's MongoDB World event in New York this week.

MongoDB Strategy and Product Marketing VP Kelly Stirman told InformationWeek in an interview that most of MongoDB is already available in the public cloud. Organizations can deploy it there themselves with an IaaS provider or on-premises and use MongoDB tools for managing the implementation. Otherwise they can use one of a few existing providers of MongoDB-as-a-service.

[What are the forces driving today's database market? Read Gartner Magic Quadrant: Open Source, Cloud Disrupt Database Market.]

The new MongoDB Atlas service provides a simpler way for organizations to consume MongoDB, because all of the back-end work is already done. Customers don't need to touch hardware provisioning, failure recovery, software patching, upgrades, configurations, or backups. They are billed by MongoDB on an hourly basis for how much they use.

"We offload all that work from your plate and let you focus on the things that let you be successful," Stirman said.

That scenario will broaden the appeal of the MongoDB platform beyond its enterprise-sized customer base, according to Stirman. It's already considered a solid platform in enterprises. Gartner named MongoDB as one of the Leaders in its most recent Magic Quadrant for Operational Database Management Systems, along with its more traditional relational brethren, Oracle, IBM, Microsoft, and SAP.

"We have over 2,000 customers, and more than half of the Fortune 100 are in production with MongoDB," he said. "That part of our business has been extremely successful. But we view cloud as an important part of our long-term strategy. We have thousands of customers who use our cloud tools to manage their MongoDB systems. This is the next step. We are going to manage it for you.

Of course, enterprise customers who still prefer to manage their own with the tools from MongoDB will continue to be able to do so, too.

MongoDB Atlas will compete against existing MongoDB-as-a-service providers. One of those is called Compose and is offered by IBM. Another one is offered by Rackspace and is called ObjectRocket.

Facebook Parse Shutdown Opportunity

The game is on between these providers and MongoDB Atlas to win over the large number of mobile developers who are currently using Facebook Parse, a MongoDB-based database service offered by the social media company that Facebook has announced will shut down in the next year.

Stirman told InformationWeek that Parse has half a million mobile apps deployed on its infrastructure. Facebook says it plans to pull the plug in January 2017. "All those apps will need a new home, and we think they will migrate to MongoDB Atlas," Stirman said. MongoDB says it believes it has advantages over competitors because of its control over the codebase of the product.

It knows about the future direction of the product. It offers features that the others don't offer including better disaster recovery guarantees. It also gives some cost advantages that may be even five times better than competitors, according to Stirman.

(InformationWeek did not independently confirm the price of each service for this article or the claims of better disaster-recovery guarantees. Pricing will depend upon the implementation.)

While the service will initially be available on the AWS public cloud, but billed by MongoDB, Stirman said that the plan is to work to integrate with the two other big public cloud providers, Google Public Cloud and Microsoft Azure. Stirman said the goal is to help customers avoid public cloud vendor lock-in.

"We'll make it easy for you to migrate between them," he said. "You are not locked into any one of those vendors at any time."

In addition to the MongoDB-as-a-service announcement, MongoDB is also using its event in New York this week to announce the MongoDB Connector for Apache Spark, a native integration that lets developers and data scientists work with real-time action on live, operational, and streaming data, the company said.

"Combining Apache Spark, the leading open source big data analytics processing engine in the Apache Software Foundation with MongoDB, the industry's fastest growing database, enables organizations to fully realize the potential of real-time analytics," said Eliot Horowitz, cofounder and CTO of MongoDB in a prepared statement.

"Spark jobs can be executed directly against operational data managed by MongoDB, without the time and expense of Extract Transform Load (ETL) processes. MongoDB can efficiently index and serve analytics results back into live operational processes, making them smarter, more contextual and responsive to events as they happen."