Follow us on social media and never miss an update
Latest Scality RING drives maximum throughput on all-flash servers
San Francisco, CA – September 14, 2020 – Scality today announced a 10x performance leap over competitive solutions with its all-flash RING for scale-out file and object storage software, and unveiled new capabilities in the latest release of its flagship RING8 platform. Scality RING customers gain the all-flash advantage with the industry’s only solution to support both scale-out file and object storage on a single platform.
RING was measured to fully saturate the bandwidth of a 40GbE network connection to a cluster of six HPE Apollo 4200 servers with 16X 12G SAS SSD drives each. RING attains this level of performance for both file (NFS; SMB) and object storage (AWS S3) interfaces equally to address the needs of legacy and new cloud-native applications.
This level of performance is the equivalent of storing or serving thousands of high-res image files each second. The result is more than ten times faster than competing scale-out storage solutions and enables the RING to be deployed across a wide range of additional file use cases that are often out of reach for other object storage vendors.
Now, a new expanded set of IO-intensive workloads will benefit from the feature set Scality customers have come to love: proven scalability, flexible choice of storage interfaces, integrated hybrid-cloud data management and the highest levels of data durability. Applications that benefit include those that previously required the use of proprietary and costly storage solutions:
- Tier-1 backup and instant restore
- Design and simulations
- Live analytics, AI & ML applications
- Single-tier medical imaging, CCTV and more
An all-flash RING deployed on a cluster of six servers can typically achieve over 10GB/s of throughput, and with performance increasing as RING cluster sizes grow, Scality customers have substantial headroom to deploy scalable systems with essentially unbounded performance as resources are added.
Modern enterprises must keep pace with a society that expects data to move at the speed of thought — always available and never down. RING has supported hybrid (flash and HDD) storage servers, with flash used to accelerate a high percentage of file system and object operations while preserving the economics of HDD for large data payloads.
Specifically for scale-out file storage (SOFS) has empowered customers to modernize with the highest levels of performance for the most data-intensive file workloads on hybrid flash/HDD, such as:
- A leading airline reservation company writes 1 PB of log data/day stored over 14 days and accessed by Splunk analytic tools, achieving 60 GB/s of NFS file write performance
- One of the largest cable broadcasters in the U.S. uses SOFS to ingest 160 high-definition video TV channels all at once
- A top European automotive manufacturer powers a 20 PB SMB share on SOFS for video data, making it one of the largest single SMB namespaces in the world
- A European national library is using SOFS to archive thousands of large digitized assets per day, scaling capacity in minutes to accommodate its quickly growing data
- A major U.S. university hospital’s PACS stores hundreds of diagnostic images daily, representing hundreds of TBs of medical image data growth each month.
“This isn’t Scality’s first rodeo with flash. We have a decade of experience using flash media in the RING architecture to accelerate a high percentage of operations for file and object access,” said Giorgio Regni, CTO and co-founder, Scality. “But now, with the declining cost curves and increasing densities of flash media, customers can use RING’s scalability and ultra-high data durability alongside the ability to fully leverage all-flash for their high-performance requirements.”
Scality’s customers have embraced RING for high-throughput use cases since 2013, leveraging the combination of flash and HDD to strike the optimum balance of price vs. performance. By leveraging all-flash storage, Scality now brings RING performance to a new level and extends its use to a wider range of application workloads.