SQL or NoSQL? In-memory or hard disks? Graph? These questions have been top of mind in recent years as developers and IT administrators check out new-age databases capable of handling scale-out data sets. Executives from four databases showed how they stand out in a hot market at GigaOM’s Structure:Data conference on Thursday.
Emil Eifrem, CEO of Neo Technology, touted the power of Neo4j and other graph databases to show relationships among disparate varieties of data with nodes, edges and key-value properties. (Think of Facebook’s Graph Search as one version.) The style takes inspiration from the connections among neurons and synapses inside the brain, Eifrem said. But, like other NoSQL databases, Neo Technology’s Neo4j product doesn’t use the SQL programming language, which could limit its adoption among enterprises.
Damian Black, CEO of SQLstream, touted his database’s use of SQL, calling it “lingua franca for data management.” Sure, it isn’t the easiest language to use. Still, “you know it’s going to save, it’s going to work,” he said. “It’s auto-optimizing. It’s proven.” Plus, it might be easier to find developers who can use it. As specialized databases get more attention, that’s become a more important point, said the moderator of the talk, GigaOM Research Analyst David Linthicum.
Different databases have different sweet spots. For Ryan Garrett, vice president of product of MemSQL, it’s comparing real-time data — from the trading floor, say — with recent historical data from perhaps a day or a week ago. Andrew Cronk, CEO of TempoDB, said his database excels at crunching time-series data in long columns coming off of many new connected devices.
Black believes storing data in memory provides a clear advantage. “It’s obviously going to be faster if you’re pulling it from memory,” he said. Eifrem took issue with that notion, saying that Neo4j runs on wherever sufficient memory is available. “Generally speaking, we want to be as horizontal as possible,” he said.
Legacy database vendors such as Oracle still command large swaths of the database market. But specialized databases such as the four on display here could keep chipping away as data sets get larger and larger. Because there are so many flavors, a few databases could become leaders, rather than just one, as they really do have different strengths and weaknesses and use-case sweet spots. At least they do for now.
Check out the rest of our Structure:Data 2013 live coverage here, and a video embed of the session follows below.
[protected-iframe id=”cb6744dc8bbf8032dd01f9f69fb26c94-14960843-25766478″ info=”http://new.livestream.com/accounts/74987/events/1927733/videos/14389771/player?autoPlay=false&height=360&mute=false&width=640″ width=”640″ height=”360″ frameborder=”0″ scrolling=”no”]
A transcription of the video follows on the next page