Companies and institutions in the bioscience, energy, government, finance, manufacturing, and other research and development sectors are increasingly realizing the significant benefits of scalable clustered computing. As cluster sizes continue to grow, scaling storage I/O performance from large numbers of clients to the same file system—or even the same file—requires deploying high performance parallel storage designed specifically to address this need.
Panasas ActiveStor 12 (PAS 12) brings plug-and-play simplicity to large-scale storage deployments and is well known for achieving blazingly fast performance levels without sacrificing reliability or manageability. PAS 12 more than doubles performance over the prior generation, scaling to 150GB/s of throughput from a single 4PB file system.
ESG documents the results of its hands-on testing of the performance scalability of PAS 12. ESG also evaluated PAS 12 with an eye on validating the claims of appliance-like simplicity in support of large scale storage environments with demanding HPC performance requirements.
According to ESG, data growth has outstripped e-mail- and database-driven structured data growth for some time now. The drivers are diverse: general purpose office data, richer file formats, ubiquitous photo and video, online communities, collaboration tools, 3-D modeling, and 4D imaging to name a few.
Economic conditions have led businesses to emphasize both CAPEX, and OPEX reductions now more than ever; file growth has resulted in higher costs in terms of additional storage, complex management, and data center floor space and energy. As a result, the ability to scale out—that is, independently scale and tune bandwidth, processing, and storage capacity on the fly while managing a single, global namespace—is extremely popular for increasing efficiency and saving money.
Adoption of scale-out storage solutions is driven by their ability to address multiple challenges. ESG survey respondents report selecting scale-out storage for a number of reasons: better scalability with easier management, improved performance of both I/O and throughput, faster storage provisioning, and reduced infrastructure cost.
But traditional file storage is only part of the challenge. Large enterprises have been using HPC to garner a competitive advantage for a long time. From technology development, to financial modeling, to seismic analysis and bio/pharmaceutical applications, HPC is critical in the development of the products and services that drive revenue in the enterprise.
HPC has a special set of requirements that traditional file storage solutions cannot address, including support for very large file systems and mount points and the ability for a large number of systems to read from and write to not only the same file share, but the same file with high throughput, to name two. Storage performance has a significant impact on HPC application performance as well: if compute nodes in an HPC cluster are waiting to access data needed for processing, the job takes longer to complete, leading to delays in time to market and potentially lost revenue.