10gen Raises $6.5M to Dominate the NoSQL Space


Updated. 10gen, the company providing commercial support for the open-source MongoDB database, has raised $6.5 million in third-round funding led by Sequoia Capital, with participation from prior investors Flybridge Capital Partners and Union Square Ventures. The money will help the company expand.

When I chatted with CEO and Co-Founder Dwight Merriman, I was more interested in how MongoDB fits into the world of NoSQL startups and how the myriad companies out there pushing other open-source database technologies could fare as they hop from customized and free deployments in web companies to building a business around paying enterprise customers. Merriman said he doesn’t believe most enterprise will have a need for a slew of customized NoSQL data technologies, but will rather end up using a combination of databases: the traditional SQL relational databases for transactions and reporting, a business intelligence data storage and warehousing product, and a NoSQL product for unstructured data.

10gen is hoping MongoDB dominates in the NoSQL category, and Merriman says with 90,000 downloads per month and considerable traction among the developer community, MongoDB is in the lead — ahead of couchDB and Cassandra. “No one is going to use eight different tools for eight different problems,” Merriman said. “The NoSQL space is one space, and popular products in that space in the long run will deliver scaling and agility of development. There won’t be 10 tools in the toolbox; it’s not that granular, and it’s just too much.”

Merriman’s confidence may be misplaced, as the jury is still out as to whether enterprises (or that many enterprises) will adopt NoSQL technologies, especially as startups attempt to tweak SQL databases to perform faster. It’s still early days, too, and there are plenty of startups raising funds to become the biggest NoSQL player on the block. (For that matter, there are plenty on the SQL side too.) For example, earlier this year Riptano raised $2.7 million to support Cassandra in the enterprise, and Membase launched (as Northscale) and scored $15 million to pursue a Memcached-based data store.

Related content from GigaOM Pro (sub req’d):



Hi, I do believe it is more or less one space, and I think you are right that my pov isn’t the conventional wisdom so far. The ‘nosql’ products do vary, for examples on these dimensions:

– data model
– consistency model
– distribution model (bigtable style partitioning vs. amazon dynamo consistent hashing)

However, many of the use cases that these products solve are in common despite the different approaches to attacking the problem. Thus I don’t think of it as different spaces, but rather different approaches to the same problem. In theory, the doc/column/kv data models all scale horizontally.

I do think of graph oriented as separate. If you have a graph problem, graph dbs can be really good. If your problem is not intrinsically a graph, I would look more widely.

Frank Mashraqi

“No one is going to use eight different tools for eight different problem”

Great! Let’s treat to solve each problem with the hammer because that’s all we got? In big data world, that’s just like shooting yourself in the foot in slow motion.


Frank Mashraqi

@Tony is right.

There are four storage formats generally in use for storage in NoSQL systems:

1. Document Oriented: document format can be binary (BSON in MongoDB) or text (JSON in CouchDB)
2. Key-Value Pairs: Most document oriented storage formats are stored as key-value pairs
3. Column Oriented
4. Graph Oriented



“The NoSQL space is one space”

That’s not true – there are 3-4 distinct kinds of NoSQL storage systems, of which Mongo fits into only 1 category.

Comments are closed.