As enterprise IT continues to move into a new age of automated systems and sophisticated data handling, Pure Storage is making headlines today for its innovative storage system.
Coverage this morning by Stephanie Condon at ZDNet shows new features for Pure Storage’s FlashBlade technology that promotes effective storage strategies for large amounts of unstructured data resources.
“The focus on real-time performance, Pure Storage says, makes FlashBlade well-suited for modern architectures, where compute and storage are disaggregated,” Condon writes. “The new version includes file replication with read-only data in the target replication site, enabling data validation and disaster recovery testing.”
Although Condon’s piece does not specify exactly how this storage system accommodates unstructured data specifically, (for example, offering detailed mining or analytics features that would use AI to aggregate insights from raw data), a separate white paper notes that the FlashBlade system “eliminate(s) storage bottlenecks for unstructured data.” In other words, much of the utility of the FlashBlade system seems to be in helping business to scale its data storage requirements.
All of those unstructured data assets, like organically generated letters and documents, take up a lot of space, and a provision like the FlashBlade vendor service can help optimize the treatment of those large data sets.
“FlashBlade is so unique because it is designed from the ground-up for tomorrow’s data pipelines and real-time analytics applications, with simplicity, ease of use and performance in mind — something that legacy storage vendors struggle with,” Pure Storage VP Matt Burr said in a press statement.
One such principal involves elastic scale-out capability for a sort of “big data transition strategy” that can be necessary as the business collects gargantuan amounts of digital data resources.
“FlashBlade is built to scale effortlessly and linearly from small deployments to very large deployments,” write Pure Storage spokespersons describing this utility. “Every dimension of the system can scale linearly with the system as it grows—IOPS performance, bandwidth, metadata performance, NVRAM and client connections.”
The authors of this service explanation also note a “low TOC” and further detail how the components of the FlashBlade system are outfitted:
“The blade is the scaling unit for FlashBlade,” notes the guide. “Each blade marries raw NAND flash with an Intel Xeon system-on-a-chip processor, a programmable processor with integrated ARM cores, DRAM and integrated NVRAM, all connected to the blade via the PCIe protocol. The Flash Translation layer inherent in standard SSDs has been re-architected to eliminate bottlenecks, and the DRAM in SSDs has been re-provisioned to significantly improve parallelism across the system.”
Look for elegant data storage architectures like these to be part of the modern company’s playbook – and factor that insight into your own portfolio research.
