I met up with DataStax earlier this week, a company building a business around the Apache-licensed Cassandra key value store. the company is very bullish about the European market for non relational database technology generally, and Cassandra specifically. One swallow does not a summer make, but a couple of swallows and the weather might just be improving. Last week I hung out with Basho, a RedMonk client, which is also building a business on a key value store in the shape of RIAK.
I found the DataStax story interesting because it confirms so many things that RedMonk has described and predicted over the years. The fact that DataStax is moving to harvest opportunities in a market already created by open source, is very different from the traditional approach to building a European beachhead. In the age of open source first, developers make the market, and then commercial firms come in after the fact to sell associated services to keep ops people happy. So DataStax claims its really fulfilling demand rather than trying to create it. The initial investment is in sales so the play is in Stockley Park.
Basho on the other hand has set up in my stomping ground, Shoreditch, Where The Developers Are. The firm has sales and technical evangelists on the ground, and serious plans to grow the team, which could double or triple in the next 12 months.
One thing I find interesting in talking to both companies is they are going after Oracle directly rather than orthogonally. And apparently customers are responsive. Another company that selling against Oracle is 10Gen, which is also investing in Shoreditch, the epicenter of the London startup scene. One homegrown entrant in Shoreditch is Acunu, a storage management company making a play as a Cassandra packager.
Demand is growing for Oracle database alternatives. Enterprises see what’s happening on the web, and want to do the same thing. Many successful web companies could not scale using Oracle because of the licensing fees, not the technical architecture. This is Oracle’s soft underbelly.
Every industry has a problem stymied by expensive relational databases- take telcos and Call Data Record analysis. Or financial services compliance- where today expectations are for 5 years of risk data to be processed every day. Before the crash 5 days was considered reasonable.
Oracle’s most recent results were not great. And now nimble well capitalised aggressive competion is going after the core franchise… not just in Silicon Valley, but in Europe, in the world of telcos and financial services. The NoSQL market is still not well understood- graphs/key value stores/document databases- but its gaining traction in the enterprise.