Startup Prophesizes Success With Cassandra Database
Cassandra–the figure in Greek mythology who told the truth about the future but was never believed–seems an unlikely name for a database. Yet the technology has believers, and Riptano hopes to create more.
- Riptano
- Riptano’s Matt Pfeil
The Austin, Tex., startup was formed to help turn the software into the next big thing for storing and analyzing vast amounts of data. Riptano this week is announcing an initial funding round, a $2.7 million infusion from Lightspeed Venture Partners, Sequoia Capital and private investor Jason Calacanis. Rackspace Hosting, another early supporter of the technology, is also an investor.
Apache Cassandra, the software’s more formal name, was originally developed at Facebook to power a feature that lets users search their inboxes. The social network then released the code in 2008 to be improved and refined by other developers, under the popular process known as open source.
That effort is overseen by the Apache Software Foundation, best known for a popular Web server program. Riptano plans to be a commercial distributor, offering support and other add-ons for companies that want help in using the software.
What’s so special about Cassandra? It is not a relational database, the variety popularized by companies like Oracle that are widely used for running high-volume financial transactions. That software tends to run best on large, powerful servers, says Matt Pfeil, Riptano’s CEO and co-founder.
Cassandra, by contrast, was designed to keep up with huge volumes of data generated at websites, borrowing from some of the best elements of databases created by Google and Amazon. The idea is to spread information across hundreds or thousands of inexpensive servers–any one of which can crash without disrupting operations. “Hardware always fails,” Pfeil says.
Users of the software include such high-volume websites as Twitter, Digg, Mahalo and Reddit. It should be noted that Twitter reversed an earlier plan to use Cassandra as the central repository of tweets, though still plans to use the software; Digg also had some technical issues with a version of its service that made use of Cassandra, though people at the company blame other factors for the problems.
No comments:
Post a Comment