NoSQL is still not well understood, as a term or a database market category, by enterprise IT decision makers. However, one NoSQL vendor, 10gen, creators of open source MongoDB, appears to be growing into enterprise accounts and distancing themselves from competitors. If you’re considering, or curious about, NoSQL databases, spend some time looking at MongoDB.
Understanding not only SQL, aka NoSQL
While the term NoSQL suggests a product category that is anti-SQL or anti-relational databases, the term has evolved to mean “not only SQL”.
According to Nosql-database.org, there are over 122 NoSQL database products to date. These products differ from traditional relational databases as they don’t rely on a relational model, are often schema free and support eventually consistent, not ACID, transactions.
While companies needing to manage terabytes of data across commodity servers, such as Facebook, FourSquare or Shutterfly have been early adopters of NoSQL databases, traditional enterprises such as Disney and Intuit are joining the NoSQL customer list.
Max Schireson, president of 10Gen, asserts that relational databases are here to stay and have an important role to play in tomorrow’s enterprise.
Schireson sees NoSQL and relational databases both being used within a given enterprise, albeit for different applications.
If this positioning sounds familiar, recall that MySQL attempted to paint a picture of co-habitation with enterprise database vendors.
If an application is processing sales orders and needs absolute guaranteed transactions, a relational database supporting ACID transactions is a must. If an application is processing millions of events, such as click streams, in order to better optimize an online sales catalog, and losing a few of those events is less critical than being able to scale the application and data across commodity servers, then a NoSQL database could be a perfect fit.
MongoDB distances itself from NoSQL alternatives
While NoSQL databases such as Cassandra, originally developed and used by Facebook, or CouchDB get a lot of media attention, MongoDB appears to be the product to catch in this hot market.
Worldwide Google searches for various NoSQL product names shows the marked increase in MongoDB and Mongo searches since January 2011. Google searches for MongoDB and Mongo exceeded searches for CouchDB, Couchbase, Membase, Cassandra, and HBase combined.
According to Indeed.com, jobs seeking MongoDB or Mongo skills have outpaced other leading NoSQL products. MongoDB and Mongo now represent the most sought after NoSQL skills amongst companies hiring on Indeed.com.
Recently announced platform as a service offerings from Red Hat and VMware featured MongoDB at the data services layer of their respective offerings.
Schireson shared some stats on 10Gen’s commercial business growth into the enterprise with MongoDB.
Six months ago the majority of 10Gen customers were startups; today the majority are traditional enterprise customers. In fact, 10Gen counts five Fortune 100 companies amongst its over 200 paying customers.
With over 100,000 downloads per month and developer attendance to MongoDB conferences increasing 400 percent to nearly 2,000 across San Francisco, New York and Beijing, MongoDB traction continues to increase.
Schireson explained that many enterprises have developers interested in MongoDB, as the above download and conference attendance data backs up. However, enterprises are waiting for their peers to go first into the world of NoSQL.
Schireson revealed that securing Disney as a public MongoDB reference has led to increased enterprise interest in 10gen from MongoDB users.
MSQL poster child Craigslist adopts MongoDB
Another recent coup for 10Gen was winning Craigslist as a customer of MongoDB.
Craigslist’s Jeremy Zawodny, author of the popular High Performance MySQL book, recently spoke about Craigslist adopting MongoDB to handle Craigslist’s multi-billion document deployment. Zawodny explains Craigslist’s evolution from being a MySQL everywhere shop to selecting the appropriate database technology based on varying data and performance needs.
When Zawodny, a MySQL performance guru, gets behind MongoDB, it’s time for enterprises interested in NoSQL to consider MongoDB.