Evaluating new hardware often starts with performance: how fast, how reliable, how much downtime. For many enterprises, performance metrics are the deciding factor between purchasing one product instead of another. Because performance traditionally translated into results, this model worked for many years.
But enterprises moving to an all-flash storage system must create a new decision matrix. With all-flash storage, performance is a given. Speed and performance are built into the core technology, meaning there is little, or no, performance variation among products.
Most Companies Will Move to All-Flash Data Centers Within Five Years
As data becomes even more integral to every part of life and business, the volume that must be stored and processed will continue to grow. As a result, IT departments must consolidate more data on less infrastructure. Because enterprises must drive value with faster and more-compelling experiences for both internal and external customers, the role of data centers will continue to expand.
To keep pace with today’s increasing data growth and app development, a new level of intelligent IT systems is necessary to tackle the data challenge and effectively bridge from data insight to action to discovery. As this trend continues, more enterprises will adopt an all-flash mindset as they undergo digital transformation. Predictive analytics, cloud-ready solutions and future-proofing investments are parts of this new wave of intelligent storage that can lead to a successful digital transformation. Enterprises must consider these new factors as they select their all-flash products.
As more enterprises have moved to all-flash storage, the challenge of evaluating all-flash storage products has become more pronounced. A recent TechTarget Research survey found that 51 percent of storage customers anticipate they’ll operate an all-flash data center within the next five years.
This result means we’re moving into the third wave of all-flash storage, where this technology is the new normal. Performance (the first wave) and economics (the second wave) are no longer accurate measures for finding the best all-flash product for enterprises. It’s time to create a new decision matrix for all-flash storage, which is quickly becoming the default storage tier for business-critical applications.
Instead of focusing heavily on performance, enterprises should consider the following three new factors as they make their product choices.
Factor 1: Predictive
Enterprise customers and employees expect data to be always available and, more importantly, instantly available. When data doesn’t exist or doesn’t provide the right information, an app data gap exists. By using artificial-intelligence tools for predictive analytics in the data center, enterprises can collect the large breadth and depth of data for the real-time analytics people expect today—and doing so closes the gap.
All-flash storage pulls telemetry data from the infrastructure stack, giving a complete cross-stack view. Using all-flash storage with predictive capabilities means enterprises can use the data to anticipate and prevent issues across the infrastructure stack. By improving how infrastructure is managed and supported, they can also increase productivity and efficiency.
Enterprises can also use machine learning and predictive analytics in the data center to provide a more personal customer experience. Real-time predictive analytics can help solve customer problems and prevent additional customer-service calls. Given the increasingly competitive market in many industries, excellent customer service is one of the few true differentiators between competing products.
Factor 2: Cloud-Ready Storage
Enterprises have moved to cloud storage because of its agility and cost benefits after having experienced increasing costs and siloed data with legacy infrastructure. While running the day-to-day IT needs of the enterprise, IT departments are also creating the strategy for taking advantage of cloud-based offerings. Although many enterprises use cloud storage primarily for long-term retention and disaster recovery, new uses are emerging, especially in cloud applications and cloud-app development in the data center.
When considering storage options, it’s essential to keep in mind that data doesn’t just stay in the data center. It must easily move between on-premises devices as well as the public cloud. Regardless of their current use of the cloud, enterprises must consider being cloud ready as an important requirement when evaluating storage solutions. With cloud-ready storage, they can deploy flash everywhere and assign applications to the best type of storage to match the price and performance requirements. As needs change, the assignments can be changed without disrupting the business.
Enterprises that aren’t using the public cloud have the confidence that they can quickly move to the public cloud at any point, creating agility in the data center. Modern approaches to storage can natively link primary storage, secondary storage and public-cloud storage, so enterprises can move application data easily and avoid lock-in. For example, they can host primary volumes and archive tiers in Azure and AWS. On-premises storage deployments are therefore cloud ready as soon as they’re deployed.
Factor 3: Timeless
Not long ago, IT departments purchased a storage system, used it for three to five years and then went through a painful rip-and-replace upgrade to keep up with growth. Now, with the performance and density of today’s flash systems, refresh cycles are getting longer. Other technology interacts with data centers, however—especially processors and media technologies, which still have a relatively short refresh period.
Enterprises often perform tasks needing instant results, such as real-time analytics, business intelligence and real-time trading. These applications increase performance demands and the need for low-latency storage. But we don’t know exactly what tomorrow’s technologies will require. In recent years, this fact0 has meant costly replacements and upgrades on relatively new infrastructure.
Enterprises are also moving away from large capital expenditures and a capex model. Many are now moving to an opex IT model, which means using a consumption model that eases accounting and administration. When enterprises use all-flash storage with a cloud-like pay-as-you-go model, costs go down and ROI increases. Additionally, they only pay for what they use and avoid large expenses to depreciate when equipment becomes out of date in a few years.
The solution involves looking for infrastructure that provides a path to the next technology. When purchasing all-flash storage, enterprises must look for systems that can scale to new technology not yet invented. They should look for vendors with reliability guarantees and built-in tech refreshes. With future-proof platforms and guaranteed availability, enterprises have much greater architectural and financial flexibility.
It’s easy to focus on performance and affordability—they’re familiar and quantifiable metrics to use when evaluating all-flash storage. But focusing on these qualities risks missing out on the capabilities and intelligence needed to keep your enterprise evolving with technology. It’s time to move past these antiqued measures and evaluate all-flash technology on the basis of the criteria that matter today and will matter tomorrow: predictive, cloud-ready and timeless.
About the Author
Milan Shetti is Storage Division General Manager of Hewlett Packard Enterprise.