AtScale is easing the way for enterprises to run big data business intelligence and analytics directly on their Hadoop clusters.
During the Strata + Hadoop World 2016 conference in today, the San Mateo, Calif.-based big data analytics software provider announced the release of AtScale Intelligence Platform 4. This latest version solves one of the biggest roadblocks faced by enterprises seeking to link their Hadoop clusters with their business intelligence (BI) tools, according to Bruno Aziza, chief marketing officer for AtScale.
"If I give you access to a Hadoop cluster, I give you access to the whole data set," he told eWEEK. Clearly, this is a "showstopper" for IT administrators at security-conscious companies and organizations that operate in regulated industries. Few banks, retailers or healthcare organizations are willing to risk giving users completely free rein over their customer and transaction data for the purposes of extracting some business insights.
AtScale 4.0 ships with the company's True Delegation software technology. It supports Microsoft Active Directory, Lightweight Directory Access Protocol (LDAP) and Kerberos and integrates with the Apache Sentry security framework and Apache Ranger role-based authorization module for Hadoop, to ensure that users are authorized perform queries on a Hadoop cluster.
The updated software includes role-based access controls that automatically sync with Active Directory or LDAP security group. The capabilities help organizations finally bring their Hadoop-compatible BI tools in compliance with their data governance requirements, potentially increasing adoption, said Aziza. Combined, the new capabilities provide "the same level of security that you expect from an enterprise security platform," he added.
Also new is Hybrid Query Service, which supports both the MDX and SQL query languages. According to Aziza, the typical enterprise uses five to 10 different BI tools from the venerable Excel spreadsheet software from Microsoft, which uses MDX, to Tableau, which favors SQL, and dozens in between.
The technology natively supports both MDX and SQL, enabling practically plug-and-play support for nearly any business analytics client software without having to download custom drivers. "Hybrid Query Service allows us to interact with any business intelligence tool in the ecosystem today," asserted Aziza.
Noting that AtScale doesn't offer its own BI client apps—although the company does offer a companion app for browsing cubes and other—the AtScale's platform helps prevent vendor lock-in. "We act as a middle layer that works with any BI tool," said Aziza. "The competition forces you to use theirs."
To accelerate queries, AtScale employs a real-time machine learning engine. Although highly capable as a big data platform, Hadoop is optimized for storage note interactive queries, reminded Aziza.
To improve performance, AtScale analyzes queries that are made by users and continually optimizes its adaptive cache. Essentially, the more users seek out new insights, the more predictive and faster the system gets. Initial queries may take a few seconds, but eventually AtScale will help deliver answers in a second or less, providing a responsive user experience for BI apps, he said.