Does the rise of cloud computing mean that traditional structured query language (SQL) databases and solutions are dead, or dying? Not according to a panel of database companies at the GigaOM Structure conference. Although the “noSQL” movement is gaining steam, most of the panelists agreed that there is still a place for traditional SQL solutions, but they need to adapt in order to provide the kind of scalability that cloud-based services and products require.
Clustrix founder and CEO Paul Mikesell said that existing SQL solutions can’t scale in the cloud because they’re built to be distributed single-box solutions, but that his company and other startups are trying to take those SQL databases and make them applicable in a horizontally scalable format so that they work better with cloud applications and services. And Mike Hoskins, chief technology officer of Pervasive Software, said that while companies may be moving away from SQL databases, the technology will have to survive for some time because so many companies use it for their critical data.
What is changing, Hoskins said, is the “death grip” that SQL and relational databases have had on all corporate data solutions. That’s ending, he said — not just in the cloud but everywhere, as more companies are looking for massively scalable solutions. Emil Eifrem, CEO of Neo Technology, said that from the development side, the reality is that “most developers hate SQL and have spent years trying to get away from it,” and that is what has given the noSQL movement so much strength. Roger Bodamer of 10gen, developers of the noSQL solution Mongo DB, agreed that the “big elephant in the room” for relational database systems is that in many cases they are extremely difficult to use.
Mikesell said the key point for most companies that are looking for cloud solutions is scalability. “That’s what everybody’s after,” he said. “The question is, does that mean you have to throw out SQL?” The Clustrix CEO said that the problem is when you throw out all that relational capability, “you end up having to implement a lot of that in your apps.” According to Mikesell, it isn’t that the SQL paradigm is broken, it’s that the implementation of those solutions has failed to scale properly.