Qumulo update, Quantum arrays target unstructured data
File vendors Qumulo and Quantum product launches differ in approach, but the impetus is identical: giving enterprise data centers the tools to subdue sprawling unstructured data.
NAS vendors Qumulo and Quantum separately launched storage products this week that provide distinctly different methods to tame unstructured data growth.
Qumulo expanded its Cloud Q configuration templates for scale-out storage in Amazon Web Services (AWS). Rebounding vendor Quantum Corp. introduced the H2000 series of hybrid arrays, which provide on-premises storage for media production workflows.
Unstructured data includes multimedia, streaming video, IoT and other data applications that don't adhere to a hierarchical file format. That's led NAS vendors to adopt scale-out architecture that treats data as objects.
Qumulo piggybacks on AWS cost reductions
Qumulo Cloud Q is used to build petabyte-scale data lakes in the public cloud, either directly from Qumulo or the Amazon Marketplace. For the most part, cloud providers have not made it easy to support scalable high-performance file storage in the cloud, said Steve McDowell, a senior analyst at Moor Insights & Strategy.
"In the public cloud, you're largely left to assemble your own solution from a set of building blocks. What Qumulo delivers is a set of technologies that harness AWS-provided building blocks and turns them into a holistic data management environment," McDowell said. "Qumulo is providing a core set of technologies that give you a consistent experience between clouds and on-prem environments. That gives an IT storage administrator flexibility to put data where it best fits."
That flexibility is important at Industrial Brothers, which uses Cloud Q-configured storage with AWS compute to avoid bottlenecks when rendering images in the cloud. The Toronto-based studio produces animated children's content, including the Universal Kids-produced Remy & Boo series.
Jean-Paul Godmaire, who heads Industrial Brothers' IT, said the expanded Cloud Q options give him the ability to flexibly adjust for scale and cost. The studio runs Qumulo file storage for high-performance and remote AWS Spot Fleet GPU instances.
"It's not sustainable for us to keep a cluster running full-time in the cloud. We need the ability to burst into the cloud and sync the data we need. That allows us to buy only the amount of Qumulo storage to get the job done. With these new configurations, we can better target the amount of data we require and order the exact formula to get the [capacity] we need for the time we need it," Godmaire said.
Cloud Q takes advantage of AWS cost reductions for its Elastic Block Store, said Barry Russell, the senior VP of Qumulo's cloud business. Russell said the expanded options give customers "maximum scalability, up or down" while lowering file-based cloud storage up to 70%.
New Quantum array for unstructured data
Quantum's flagship is the StorNext File System, which typically runs in a metadata controller. Users scale StorNext storage by adding block storage devices. The H-Series hybrid array is nearline storage based on SAS SSDs and HDDs. It is intended for unstructured data workloads and designed to complement Quantum F-Series NVMe flash arrays for tier 1 storage.
The arrays support 32G Fibre Channel, 100 Gigabit Ethernet connectivity, and an AMD Epyc-based PCI Express Gen 4 architecture. Two H-Series models are available: the 2U H2012 that scales to 160 TB of usable capacity per chassis, and the H2024 that is rated at 307 TB after data reduction. H-Series systems are available now, with a list price starting at $38,000.
About the author
Before joining TechTarget's news team, Garry spent more than 13 years as a freelance business and technology writer, including 10 years as a contributor to TechTarget. Garry's work has appeared on SearchNetworking, SearchSecurity, SearchMicroservices and other TechTarget sites. Prior to freelancing, he was a staff editor at Digital Business and managing editor of Richmond Ventures magazine.