ebay CEO John Donahoe

The past week has seen not one but two technology giants take on a bigger role in the open-source community’s efforts to bring modern analytics within the grasp of the traditional enterprise. Online retail giant eBay Inc. led the charge with the contribution of a homegrown OLAP engine that it said makes querying Hadoop both easier and significantly faster.

Dubbed Kylin, the technology has a back story that’s familiar to the Hadoop community: A web company needed to address a new problem at a scale no existing product could handle, so it built its own. The particular challenge eBay faced was sustaining the response times its business users had come to expect for queries as the amount of information outgrew the capacity of its environment to handle it effectively.

Kylin solves that issue by automatically mapping certain repeating fields in HBase, which the e-commerce powerhouse runs on top of Hadoop as a data warehouse, to the complementary Hive component in the form of key-value pairs. That eliminates the need to sequentially scan upwards of billions of rows in HBase every time a business worker needs a more complicated question answered about about customers or products.

Et tu, Oracle?

Continuing its love-affair with open-source software, Oracle Corp. followed suit with the introduction of a new platform for tracking the flow of information inside Hadoop and other popular solutions that aims to bring the same kind of nimbleness to the data integration phase, which is another historic bottleneck. The Enterprise Metadata Management 12c toolkit provides an environment for analysts to extract insights about how data travels through their organizations that can help with spotting everything from innocent input errors to malicious corruption of sensitive business information.

Yet while it marks another milestone in the database market’s cautious warm-up to open software, the launch also serves the almost contradicting goal of strengthening Oracle’s grasp over its technology ecosystem, which is the same objective Teradata Corp. is aiming for with Connective Analytics. Introduced shortly after the Metadata Management suite, the offering extends the company’s value proposition beyond data warehousing with capabilities for mapping relationships across various online entities such as social media accounts and customers. According to the firm, the software makes it easier to identify complicated patterns that might otherwise slip through the cracks.