Sure the headline seems a little dramatic, but we’re not exaggerating at all.
Consider this from the pitch for coverage we got from Couchbase last week:
"Couchbase will announce Server 2.5 to support enterprises as they transition away from experimentation with first-generation NoSQL developer projects, such as MongoDB, to the development and deployment of a true enterprise-class NoSQL product to scale mission-critical applications.”
For anyone who needs a translation, or at least our take on the statement, here’s the skinny:
Couchbase claims that:
- Enterprises are transitioning away from “first generation” NoSQL-based solutions (such as MongoDB)
- MongoDB (and other first generation NoSQL products) has been used by developers for “experimentation”
- Mongo DB isn’t a “true” enterprise class NoSQL product, it can’t scale for mission-critical applications
- Couchbase Server 2.5 is the product of choice as enterprises transition from NoSQL experimentation (with products like MongoDB) to getting serious about NoSQL for enterprise-grade mission-critical applications
Did We Get That Right?
We don’t usually receive coverage pitches and press releases seeded with digs at competitors (though Larry Ellison does occasionally make them in front of live audiences), so when we interviewed Couchbase CEO Bob Wiederhold we asked him a pretty blunt question, “Are you saying that you’re going after MongoDB?”
He didn’t back off.
“I am saying that we’re going after MongoDB’s customers,” he said
What is Couchbase Server 2.5, Anyway?
“Couchbase Server 2.5 was built for the enterprise,” said Wiederhold. He explained that the document-oriented database features high availability and rack zone awareness -- it stores master data and replicated data on different server racks insuring that data remains highly available and secure even if a server goes down. Not only that, he added, it’s also easy to scale and administer.
Speed and low latency are also critical in this day and age, according to Wiederhold, and, as such, Couchbase Server 2.5 features Secure Cross Data Center Replication (XDCR).
“When data is not stored on virtual private networks (VPNs), it’s usually sent between networks over wide area network,” Wiederhold explained. “The problem with this is that data can become vulnerable unless it’s encrypted."
Couchbase addresses that problem with SSL transmission between data centers. The new security layer helps enterprises achieve global scale in a secure manner, according to the company.
Is Couchbase Server 2.5 Better for Enterprises than MongoDB?
Couchbase and MongoDB could probably go feature-by-feature and toe-to-toe in answering that question, so we’re not going there.
It is interesting to note, however, that Couchbase recently won a customer, Viber, from MongoDB. Viber, one of the fastest growing messaging and VoIP services in the world and driver of billions of messages every month, said that MongoDB couldn’t scale.
This from a Couchbase press release:
We found that MongoDB’s NoSQL technology could not adequately support the company’s growing data management needs,” said Amir Ish-Shalom, systems architect at Viber. “After investigating multiple database technologies such as Cassandra and Amazon’s DynamoDB, including more combinations of database plus caching layers, we deployed Couchbase. With Couchbase, we’re now able to handle hundreds of thousands of ops-per-second -- an order of magnitude more than our previous solution.”
While Couchbase points to this as proof of its superiority over other NoSQL technologies, like MongoDB (and it’s not out of the realm of possibility), we don’t consider a single win as representative of a trend.
The fact that Couchbase grew 400 percent in 2013 is impressive, but this could simply speak to unprecedented interest in NoSQL technologies. MongoDB, for example, has enjoyed skyrocketing growth; it just surpassed IBM DB2 to become the fifth most popular database in the world.
There are four databases that are more popular than MongoDB at the moment: Oracle, MySQL, Microsoft SQL Server and PostgreSQL. Why not go after them we ask?
Care to Comment MongoDB and DataStax (Cassandra)?
Due diligence requires that we ask other leading NoSQL vendors, MongoDB and DataStax (Cassandra) for their comments. Here’s what they said:
Matt Asay, VP, marketing, business development & corporate strategy at MongoDB:
We never comment on other vendors’ news releases, and we do a pretty great job of not commenting on competitors at all. We're not here to beat this or that competitor. That's weak. We're here to enable engineers to build and run modern applications. We care about customers, not competitors, and are enabling the next generation of developers to be far more productive than legacy relational database technology allows."
Matt Pfeil, co-founder of DataStax:
In an online environment, availability matters more than anything. 99.999% uptime doesn’t cut it anymore -- 100% is the new requirement to do business. A new age of Big Data requires new approaches to data management. Traditional systems such as Oracle cannot meet the needs for always-on availability in a data rich environment, and that’s exactly the promise that Cassandra delivers.
Cassandra was designed in 2008 with Big Data in mind, and it utilizes advanced technology (such as peer-to-peer replication) to guarantee always-on availability even when failure strikes. No other database, including Oracle, Couchbase or other NoSQL providers can deliver always-on availability like Cassandra because of its core architecture."
Focus On Solving Problems
As time passes, legacy databases will outgrow their usefulness, practicality and viability, leaving NoSQL vendors with a world of opportunity before them.