What to Look for in an All-Flash Array
It’s that time of year when IT industry analysts make predictions about the trends that will drive technology spending. Storage remains a major line item on 2018 IT budgets, and more and more organizations are investing in flash storage platforms as opposed to legacy, disk-based systems. According to the Spiceworks 2018 State of IT report, flash will become the primary storage for many workloads this year, with disk drives moving to secondary storage and archival applications alongside magnetic tape.
The need for improved performance is the primary driver of flash adoption. Hard drives are mechanical, and the time it takes to move the read/write arm across a spinning disk creates an unacceptable amount of latency in today’s data center environment. Solid-state flash drives are electronic, with no moving parts. As a result, flash storage is more than 20 times faster than disk, with sub-millisecond data access times.
All-flash arrays are also smaller than traditional disk drives and use a fraction of the power. Experts say that reduced power and cooling costs can more than justify investments in flash storage. Furthermore, many organizations find that all-flash arrays give them enough performance and capacity to consolidate workloads, enabling them to reduce the number of storage arrays that must be purchased and managed. In addition, most flash arrays incorporate data optimization techniques such as de-duplication and compression to dramatically increase capacity.
However, many all-flash arrays are built using legacy architectures that were designed to increase the performance, availability and data protection of disk storage. These architectures are not capable of taking full advantage of the high performance and low latency of flash – in fact, flash can overwhelm the controllers, ports, I/O interconnects and other components. At the same time, it’s important to recognize that flash is less reliable than disk, statistically speaking. Not all flash storage systems meet enterprise requirements for availability and resilience.
Flash-optimized architectures are designed for the performance characteristics of flash. Every component of the system is optimized for flash, ensuring that workloads have all available resources at their disposal when needed. These solutions are scalable, and provide consistent, predictable performance. They also provide enterprise-class data protection and high levels of reliability.
HPE Nimble All-Flash Arrays deliver enterprise-class features in a flash-optimized architecture. With sub-millisecond response times and six-nines guaranteed availability, HPE Nimble All-Flash Arrays can support the most demanding, mission-critical workloads. Variable block de-duplication and compression provide five times the data reduction of competitive products with no performance impact. HPE Nimble All-Flash Arrays are simple to deploy, configure and manage, enabling you to scale up performance and capacity independently and non-disruptively, and scale out to four arrays managed as one.
The HPE Nimble Storage Predictive Flash platform also includes an advanced predictive analytics engine for organizations seeking a true next-generation all-flash solution. HPE InfoSight predictive analytics is capable of addressing over 86 percent of storage issues before they occur. This technology offers both superior performance and the highest availability with options to scale to petabytes to meet growing requirements.
If flash is on your radar for 2018, look for purpose-built solutions that deliver the performance and reliability your organization demands.