[Video] A Sneak Peek Under the Hood of WANdisco’s Big Data Software

WANdisco’s Jagane Sundar and Brett Rudenstein showcased their firm’s new software at Hadoop Summit 2013. SiliconAngle NewsDesk Head Writer Winston Edmonton stopped by their booth to take a look.

Winston inquires about WDD 3.6, WANdisco’s S3-enabled distro of the Hadoop Distributed File System (HDFS). Sundar tells him that the platform implements the S3 API, which means that users can migrate their apps from Amazon Web Services service to the private cloud with minimal effort. He highlights that storing data in an on-premise Hadoop deployment is ‘orders of a magnitude’ cheaper than alternatives such as AWS, and adds that bringing workloads in-house has the added benefit of addressing security concerns associated with public cloud services.

After going over WWD 3.6, Sundar and Rudenstein show Winston Non-Stop NameNode WAN Edition, a utility that leverages WANdisco’s patented active-active replication
technology to eliminate the single point of failure in geographically dispersed Hadoop environments. Sundar says that the system maintains not one but several NameNodes, or Hadoop metadata servers, to guarantee continuous uptime in the event of a failure. He points out that the platform also simplifies disaster recovery by synchronizing HDFS across multiple facilities. This functionality nullifies the chance of data loss and split-brain scenarios in situations where an entire data center goes offline.

Sundar boasts that several financial services companies and healthcare providers expressed interest in deploying Non-Stop NameNode. He explains that firms in regulated industries have to meet strict compliance requirements for safeguarding large volumes of data, the very task WANdisco’s platform is designed to accomplish.

Click the video below for the full segment, including a demonstration of Non-Stop NameNode’s ability to replicate data across multiple facilities regardless of distance.