Spotting an emerging market for cloud-based Hadoop, Silicon Valley startup Altiscale has officially launched Altiscale Data Cloud, a Hadoop-as-a-Service (HaaS) offering the company says is designed to free users from the complexities of deploying, managing, and scaling a big data platform.
In a phone interview with InformationWeek, Altiscale CEO and cofounder Raymie Stata, formerly chief technical officer of Yahoo, said the new service helps simplify Hadoop for companies that lack the expertise or resources to run the big data framework without outside help.
"We have the world's only purpose-built Hadoop cloud," said Stata. "It truly is Hadoop-as-a-service -- not Hadoop as yet another application slapped on top of a generic infrastructure-as-a-service."
[Too many companies fall back on incremental fixes. Here's how to foster creative change. Read Why Most Companies Renovate Instead Of Innovate.]
At big companies like Yahoo and Google, large teams of well-trained experts keep Hadoop -- or a similar big data platform -- up and running. "It's run by a central team that not only keeps it running like clockwork, but also keeps it expanding because it's always growing," Stata said. This group also keeps Hadoop up to date, he added, because the software's changing a lot.
Altiscale Data Cloud, however, is targeted at businesses that lack the technical staff and infrastructure to manage Hadoop on their own.
After founding Altiscale in 2011, Stata and his colleagues looked around at other Hadoop installations in the wild and found many of them lacking. "What we found was, they really had subpar installations," said Stata. "They were quite small -- five, twenty, fifty nodes, maybe. They were run by people who maybe [spent] a quarter of their time on Hadoop, but who had a lot of other stuff to do."
This wasn't good news for business users hoping to glean actionable insights from those Hadoop installations. "They didn't have a deep operations team," Stata said. "As a result, the folks on top … were really kind of on their own in terms of getting the most out of Hadoop."
The Altiscale founders spotted a market niche: Bringing a Yahoo-style Hadoop experience to business with fewer resources. "We saw an opportunity," said Stata. "Let's bring that awesome kind of Ferrari of Hadoop to these users and give them a pit crew to help them navigate the track and win the race."
The pit crew in this case was Altiscale's team of Hadoop veterans. "We have a lot of Hadoop expertise, so we provide a lot of support for users to make them more effective at using Hadoop," Stata explained. Altiscale's ability to install, manage, and update a business's Hadoop installation is a key distinction between it and other cloud-based providers, he added.
The company offers the Hadoop Distributed File System (HDFS) for storage. "As our customer, you get an HDFS volume, and you can store stuff in it for as long as you want -- from 10 terabytes to petabytes."
"Behind the scenes, there's what we call auto-elasticity: If you need more capacity because you're storing more stuff or running bigger jobs, we take care of that for you," he said. "And we just charge you for the resources you're using, the bytes you're storing, and the compute time you're consuming."
The HaaS market is still largely untapped, as the company has only about ten customers at this point. "They are mostly digital media, digital advertising, and SaaS companies -- the usual Hadoop suspects," said Stata. "In terms of storage, which is a good way to size these folks, they [range] from a few tens of terabytes up to a few hundreds of terabytes."
Two Altiscale customers, he noted, aim to get to a petabyte of data by the end of the year.
InformationWeek 500 companies take a practical view of even trendy tech such as cloud, big data analytics and mobile. Read all about what they're doing in our big new special issue. Also in the InformationWeek 500 issue: A ranking of our top 250 winners; profiles of the top five companies; and 20 great ideas that you can steal. (Free registration required.)