For customers using petabytes of data in production, traditional backup and disaster recovery methods aren't feasible. One vendor is skipping the data backup step and focusing solely on recovery in the event of a cyber attack or data deletion.
Arcitecta recently added a new feature called Point in Time to its Mediaflux data management platform. Using Arcitecta's proprietary database and file system, Point in Time sits on top of the production environment and tracks inline metadata changes to provide continuous monitoring and flag suspicious activity. The metadata is compressed and stored in Arcitecta's high-performance XODB database on NVMe-based storage.
During an event, IT admins can see where the data changed and roll back the file system to a time before the event occurred, eliminating the need for a separate data backup system.
Mediaflux provides a new approach to backup and recovery products, which haven't changed much in the last 20 years, according to Phil Goodwin, analyst at IDC. While newer backup services now protect virtualized environments, few are designed for multipetabyte situations.
"Arcitecta is trying to fundamentally re-architect the way backup and recovery products operate," Goodwin said.
Arcitecta's Mediaflux Point in Time targets companies with large amounts of unstructured data. The product's goal is to have near-zero recovery point objectives and recovery time objectives by capturing changes as they occur, which minimizes data loss compared with scheduled backups that traditional methods use, Goodwin said.
Ken ClippertonLead analyst for storage, Data Center Intelligence Group
Founded in 1998 and based in Australia, Arcitecta released Mediaflux, its first product, in 2002 to collect and analyze large sets of structured and unstructured data. Now Mediaflux has added the Point in Time feature to focus on data protection for unstructured data sets, according to Jason Lohrey, founder and CEO of Arcitecta.
Backing up petabyte-scale data is not only expensive, but it's technically challenging because traditional backup and recovery are inefficient, according to Ken Clipperton, analyst at Data Center Intelligence Group (DCIG).
"Arcitecta doesn't back up the data, but it gives users the ability to recover to any point in time," Clipperton said. "More importantly, it gives users a timely recovery of multipetabyte data repositories."
Since all changes are monitored, if users have a problem, they can go back before changes occurred and begin running again, Clipperton said. This is a faster recovery compared with going to a separate backup storage repository and mounting a clean backup into production.
While this is geared toward multipetabyte environments, it could be used for smaller data protection, putting Arcitecta in competition with the traditional vendors such as Commvault, Veritas or Veeam.
Dialing in on recovery
At the heart of Mediaflux is Arcitecta's proprietary XODB object database, Lohrey said. XODB uses metadata tagging and query approaches to locate data. Arcitecta also created its own file system that sits in the data path and records all changes in XODB. The merging of the file system and XODB results in Point in Time, a feature that can scale to exabytes, even if the data is spread across different data centers and clouds.
The merging of the database and file system gives the vendor inline metadata tracking that is key to its ability to restore large volumes of data as well as determining the point in time, Goodwin said.
Arcitecta is trying to tackle one of the biggest IT issues businesses face: how to protect unstructured data, according to Goodwin.
But aside from lacking name recognition, if Arcitecta approaches the broader data protection market, customers might hesitate to move away from a traditional method of data protection, Clipperton said.
Adam Armstrong is a TechTarget Editorial news writer covering file and block storage hardware and private clouds. He previously worked at StorageReview.com.