New patent outlines high performance method for partitioning distributed databases to scale deduplication beyond the PB mark
SUNNYVALE, California ‒ January 22, 2019 — Druva, Inc., the leader in cloud data protection and management, today announced it has been issued a new patent by the U.S. Patent and Trademark Office (USPTO) for optimizing storage performance and network bandwidth usage in the cloud through its global deduplication technology. With this new patent, Druva is able to further enhance cloud-based data management and remains the only vendor able to deduplicate petabytes (PB) of data in a single index.
Today’s companies are storing and managing an increasing amount of data, upwards of multiple PBs on-premises and in the cloud. For enterprise services like disaster recovery and backup, which rely on a deduplication index to store and recall data, bandwidth and storage efficiency are critical to quickly recovering files, managing costs and meeting business SLAs. However, traditional approaches limit index size, severely handicapping the benefits to an enterprise expanding its data footprint. Only Druva’s unique and patented approach enables enterprises to effectively scale a single index beyond a PB without compromising cost or performance.
“Companies of all sizes are moving more of their data to the cloud, and data protection must be able to operate effectively and scale accordingly with each business,” said Milind Borate, co-founder and CTO, Druva. “Customers are storing multiple PBs with our service, and leveraging our unique architecture, we will be able to provide instant scalability, guarantees for performance, and ensure uninterrupted business continuity, all at a lower cost compared to traditional solutions. This new patent is further recognition of our team’s outstanding drive to deliver the industry’s ultimate Data Management-as-a-Service solution.”
Patent No. 10,133,745, “Active repartitioning in a distributed database,” was awarded to Druva in November, 2018, adding to Druva’s patent portfolio. The patent explains the repartitioning of data storage without any downtime, ensuring load-balancing of workloads in the background without interruption to users. Using active partitioning with distributed database-based index technology, Druva has created the ability to scale-out linearly beyond the PB scale without impacting bandwidth, cost efficiency or performance.