Don’t tell Oracle's Larry Ellison, but NoSQL databases are the future of the enterprise. Yesterday’s databases weren’t built to handle today’s avalanches of data streaming from social, mobile, web transactions, the Internet of Things and God knows what’s next.

And Couchbase will be the NoSQL database that enterprises who are serious about winning the future will choose, according to Bob Wiederhold, CEO of Couchbase. He said that NoSQL databases have gone through an evolution that began four or five years ago when developers discovered and downloaded technologies like MongoDB to play with at home.

“They liked the ease of development, so they brought them to work and built lightweight applications,” he added.

This was an important period in NoSQL adoption, he continued. It set the stage for 2013 when enterprises decided that NoSQL was ready for prime time and that it could be used for mission critical applications.

“This is when Couchbase took off as company,” he said. “It’s when companies like AT&T, Walmart and eBay began to rely on us, the enterprise NoSQL leader for running high performance mission critical applications."

Major Announcements

And now Couchbase is making it easier for other enterprises to do the same. Today at its annual user conference, Couchbase Connect, Wiederhold and his team will introduce a series of innovations and updates that are meant to establish how and why Couchbase stands above the competition in giving enterprises the technologies they need to thrive in the modern era. 

“We work with large customers with high demands, we push the envelope when it comes to speed,” said Ravi Mayuram, Couchbase senior vice president of engineering. He added that its new architecture, Couchbase 3.0, can handle a million operations per second, some with microsecond latency.

“It’s a game changer,” he said.

“In 2015 Couchbase will be the NoSQL database of choice when Enterprises go to re-platform,” said Wiederhold, explaining that the open source software, which his company supports commercially, offers better storage, mobility, query, and performance than its NoSQL competitors. “With Couchbase Server 3.0, we’ve massively broadened the use cases we can support.”

To prepare for this moment Couchbase has made some significant architectural changes, built-in better scalability, created a more attractive environment for developers, produced a new data store, Forest DB, and more.

All together they’ve added more than 200 new features and enhancements, including major extensions to Couchbase’s core architecture.

“This advances our lead in delivering the industry’s most scalable, reliable and consistently high-performance distributed database,” said Mayuram.

There are two first of their kind architectural innovations in the new database release: stream-based Database Change Protocol (DCP) and Dynamically Tunable Memory. Together they “dramatically” improve scalability, resilience, performance and the cost-efficiency of working with massive operational data sets.

Optimizing Performance

Dynamically Tunable Memory gives enterprises a bind time option to tune memory utilization by keeping keys and metadata in memory only for the working dataset -- rather than the full database -- thereby optimizing performance and resources for the most frequently used data.

Couchbase’s stream-based Database Change Protocol (DCP) extends Couchbase Server’s unique, memory-centric architecture for data synchronization across nodes, clusters and data centers. While other databases use a disk-based approach, Couchbase’s pure in-memory DCP removes disk IO dependencies and latency, operating hundreds of times faster than disk, according to tests that Mayuram told us about. The benefits include the speed at which data moves and monitoring for large (50-60-100 node) clusters thereby increasing the speed of business and reducing cost of ownership. Additionally, according to Wiederhold, the protocol will be exposed to third parties.

How confident is Mayuram with these innovations? “Our customers put us on their most critical systems,” he said. “We are battle-tested. We have delivered on the inevitable.”

How Does Couchbase Compare to MongoDB?

We had to ask. “They have lots of customers,” said Wiederhold, noting that it is typically used for lightweight apps. (We’d bet MongoDB has something to say about that.)

When it comes to performance and scalability, “We crush Mongo,” added Wiederhold.

And beginning today, the ease-of-development edge that MongoDB has had until now may diminish. Couchbase will announce SDKs for C, Java, .NET, Node, PHP, Python, and Ruby making it easy for developers using the most popular programming languages to build scalable and high performance next generation applications.

The 2.0 SDK releases were built not only by Couchbase engineers but also by the open source community. ”We think developers are going to be delighted by new capabilities like support for Reactive and Asynchronous programming, the ability to quickly build high performing web scale applications, and native support for JSON objects” said Mayuram.

But that’s not all, a preview of Couchbase’s 2.0 next generation query language (N1QL-pronounced “nickel”) is also being made available, giving developers the opportunity to begin to build fluency early on.

Will Couchbase Rule in 2015?

With today’s release of Couchbase 3.0 and SDK 2.0, Wiederhold and Mayuram said that their owning the enterprise SQL space is pretty much a done deal. We’d bet that the folks and communities behind MongoDB and Cassandra/DataStax would beg to differ.

At the end of the day, the best set of tools for the task at hand tends to win. And we’re not at all sure it’s a zero-sum game.

Title image from the 1953 film, The Wild One.