Altiscale Makes Hadoop Easier for SQL Pros

Altiscale, a Hadoop cloud startup founded by former Yahoo CTO Raymie Stata, has added one of the most important capabilities in today’s Hadoop market to its cloud offering – the ability to use SQL to run queries on data stored in Hadoop.

SQL is one of the most popular languages used to manage and analyze structured data, and Hadoop is one of the most popular frameworks for storing and processing unstructured data. There are a lot more people who know SQL than there are people who know Hadoop, however, which is why many businesses built around Hadoop have introduced the SQL-on-Hadoop capability.

Altiscale provides fully hosted and managed Hadoop cloud service that requires no knowledge about operating infrastructure underneath the open source framework on the user’s part, Steve Kishi, the company’s vice president of product management, said. The startup leans on this aspect heavily in trying to differentiate itself in the active Hadoop market.

“We provide Hadoop as a SaaS (Software-as-a-Service) service at the end of the day,” Kishi said. “Our customers don’t have to think about infrastructure.”

Competitors, such as Amazon Web Services’ Elastic MapReduce, require users to install and operate Hadoop on cloud infrastructure.

Another Hadoop cloud startup, Qubole, offers its service on top of the AWS cloud but can also set up on-premise Hadoop clusters with elastic capacity provisioning at customers’ own data centers.

Altiscale has its own data center infrastructure to run the service on, pitching it as Hadoop out of the box. “That’s pretty critical in this space because of the complexities of Hadoop,” Kishi said.

He describes the startup’s billing model as similar to a modern cellphone subscription plan. Customers pay for resources they use.

If you signed for 10 TB of HDFS storage capacity and 10,000 task hours of compute, for example, you would pay about $2,500 per month, Kishi said.

Data center provider as channel partner

Altiscale uses colocation providers to host its infrastructure in one facility on the east coast and two on the west coast. The company has a symbiotic relationship with Carpathia Hosting, its east coast data center provider, which has referred some of its colocation customers to it.

“We’re finding that data centers can be a very good channel for us,” Kishi said. “We’re finding it to be very profitable.”

Some of Altiscale’s bigger customer wins came as a result of Carpathia referrals. Carpathia benefits by offering its customers a Hadoop-as-a-Service provider whose infrastructure they can connect to directly in the data center.

“We can set up direct connects between racks inside the data center and transfer data more efficiently that,” Kishi said.