Three months after introducing Linux support, WANdisco is unleashing an S3-enabled version of HDPS that brings public cloud functionality behind the firewall.

WANdisco Distro (WDD) 3.6 supports all the features of Amazon’s popular file storage service, including searchable metadata tagging, snapshots, and data object versioning. The platform is compatible with the S3 API and packs a set of utilities that let users import data from webserver logs, data appliances, RDBMS servers and other traditional silos.

“Feedback from the Wikibon community indicates that the public cloud, specifically the Amazon Web Services cloud, is a popular choice for experimenting with Hadoop and Big Data application development,” said Jeff Kelly, Wikibon’s Big Data Analyst. “Many enterprises, however, prefer to bring Hadoop back in-house when moving to production. WANdisco Distro with S3-enabled HDFS makes migrations significantly simpler, allowing enterprises to seamlessly move data from the Amazon cloud to in-house private cloud environments without needing to rewrite Big Data applications.”

WDD 3.6 is available for free download. WANdisco will release interoperable versions of Spark and Shark, an ultra-fast MapReduce alternative and the associated data warehousing system, within the next couple of months.

The vendor is complementing its open-source Big Data lineup with Non-Stop NameNode WAN Edition, a network solution that applies active-active replication to distributed Hadoop clusters. By turning every NameNode into a synchronized mirror of other metadata servers, the software delivers high-availability and eliminates the need for load-balances and NAS devices. It also avoids so-called split-brain scenarios where data is corrupted due to scope overlap.

Jeff Kelly explains that Non-Stop NameNode resolves the single point of failure that is inherent in Hadoop. This makes the system viable for enterprises that require “continuous and uninterrupted cluster availability.”