We are in the third platform computing era and enterprises of all sizes are being asked to continue to support legacy applications while providing support for next-generation applications in mobile computing, social media, big data/analytics, and cloud environments. As most IT organizations are consolidating around virtual infrastructure for both legacy and new applications, this has significantly changed the I/O profiles that storage infrastructure has to deal with.
It has become clear that for more primary storage workloads, hard disk drives (HDDs) no longer adequately meet the requirements of IT organizations. As a result, enterprises are shifting their storage revenue away from HDD-based designs toward newer flash-based designs. Flash-based designs offer many benefits including:
Flash-based designs are rapidly coming to dominate primary workloads as a persistent storage medium.
For latency-sensitive workloads, flash-based designs offer performance, capacity, storage density, energy, floor space consumption, efficiency, and reliability benefits that are better than the benefits from systems based around HDDs. All-flash array (AFA) platforms are being considered by more and more enterprises as they can handle third platform computing workloads, unlike HDDs.
Differing workload requirements are driving the rise of new storage architecture.
In the past, most enterprise storage designs were based around a scale-up, dual-controller model but over the past few years, new architectures have emerged including scale-out, software-defined, converged and hyper-converged infrastructure, and cloud designs. These new architectures offer benefits for many next-generation application (NGA) workloads. These new architectures offer the ability for enterprises to scale-up, something dual-controller designs could not offer. All-flash versions of these new architectures are now available from enterprise storage providers and they continue to meet increasingly stringent performance requirements.
Decreasing flash costs are enabling its use in workloads that aren’t just about low latency.
Traditionally, flash-based designs have been used with mission-critical, latency-sensitive primary workloads where better performance translated directly to improved bottom-line business benefits. However, as flash-based design costs have continued to decrease, the use of these designs with less latency-sensitive workloads has emerged as a market. This has significantly increased the total available market that vendors can address with their all-flash products, allowing them to expand the use of flash-based designs to unstructured workloads, as well as the traditional block-based workloads.
As more and more primary workloads move to all-flash storage configurations, vendors that offer a broad portfolio of all-flash options, targeted for different customers and workloads will find themselves in the best position to meet today’s enterprise storage requirements. As the all-flash market continues to evolve and meet the different needs of customers, there are unique system differentiators and workload targets for each platform.
Highly scalable, truly enterprise-class all-flash platforms are finally here and they’re being used and considered as a general-purpose platform for mixed primary workload consolidation. Many IT organizations are beginning to evaluate their legacy storage platforms and considering AFAs as a replacement. Flash-based design costs continue to drop and more and more enterprises are moving to all-flash for primary storage solutions.
Flash-based designs not only offer high density, low energy, and minimal floor space consumption but they’re also highly reliable and able to suit the variety of different storage needs that enterprises have. If your enterprise doesn’t understand How Broad All-Flash Vendor Portfolios Help IT Customers, this whitepaper can help.