Data integrator Informatica and fledgling Apache Hadoop developer Hortonworks jointly announced Nov. 2 that Informatica's HParser is now available free of charge to Hortonworks Data Platform users and available for download from the Hortonworks Website.
The data parser is also included as part of Informatica's freely downloadable Community Edition. Hortonworks is a Hadoop development spin-out from Yahoo that was launched last July at the annual Hadoop conference.
Big data workloads often contain numerous file formats, not all of which can be processed-or processed efficiently-by the Hadoop engine, which is designed to handle large batches of data. The HParser enables users "to extract meaning from data in any encoding and format, then process it to look for analytic patterns, or store it, and so on," Informatica Senior Vice President and General Manager Juan Carlos Soto told eWEEK.
"That step is typically considered parsing. It is then put into a format that is easier for you to process going forward."
Runs on Most Distributions of Hadoop
Informatica's HParser runs on virtually any distribution of Apache Hadoop. It uses the parallelism of the MapReduce framework to efficiently turn unstructured complex data, such as Web logs, social media data, call detail records and other data formats, into a structured or semi-structured format in Hadoop.
Once transformed into a more structured format, the data can be more rapidly used and validated to drive business insights and improve operations, Soto said.
HParser Community Edition provides users with an interactive tool to simplify and speed the data parsing and analytics process for some of the most often-used data types in Apache Hadoop, such as logs, Omniture, XML and JSON.
"Informatica is taking a critical step toward making Hadoop easier to use together with Hortonworks, a team that has contributed more than 80 percent of all the code in Apache Hadoop," Soto said. "We are empowering enterprises to turn big data into competitive advantage and ultimately maximize their 'return on data.'"
The HParser Community Edition includes a visual, drag-and-drop utility for quickly and easily creating parsing definitions.
Background on Hadoop
Apache Hadoop, open-source software built in Java that works with distributed data-intensive applications, enables applications to scale securely in order to handle thousands of nodes and petabytes of data. A number of companies now are using Hadoop daily to predict business patterns, find tendencies in scientific data and predict the weather, among many other functions. More and more businesses are finding out that they need to analyze their stored data and use those metrics to help them make better business decisions. Hadoop has certainly created the most buzz among new-generation big-data analytics packages.
Hadoop has gone from creator Doug Cutting's science project to mainstream business in a short time. Hortonworks was created during the last four months as an independent, privately held, VC-funded company to lead the Hadoop community and market the open-source product into the future. Officially, Mothership Yahoo is now one of its customers.
Hortonworks is an appropriate name for the new company because it is congruent with Hadoop itself-which is named after the stuffed toy elephant that belongs to Cutting's young son.