Making the Case for All-Flash Storage

By Travis Foschini, Solutions Architect 

Many of the IT managers we work with are considering all-flash arrays in their on-premises or colocated data centers. While the cost per GB for flash storage is coming down, it’s often considered a luxury because of the way storage costs are traditionally evaluated. Despite the many benefits of flash, it can be difficult to gain approval for sufficient funding from the fiscally-minded business executives. In this post, I’ll lay out four primary arguments that can help get you over this hurdle.

AdobeStock_puzzle_2.jpg

4 Reasons All-Flash Storage Makes Sense

The case for flash storage is both a technical and a financial one. Both can be useful for convincing the powers that be.

#1 Performance — While many executives are concerned about bottom-line costs, we find the performance argument is often enough to get them to refocus their priorities. Over the years, drives have grown larger and larger. You can now buy individual spinning disks with capacities of 10 Terabytes or more. That’s great if all you care about is storage capacity, but if your organization is making regular updates to data or performing queries on that data, it can be a problem. That’s because as spinning drives get larger, the performance characteristics of each disk remain the same.

#2 Cost — We just touched on cost in our performance argument, but there is another cost element (also related to performance) that can strengthen your case. With traditional spinning disk drives, improving performance usually requires buying more drives. It’s far too easy for a business that takes that route to throw good money after bad as they try to bring performance up to expectations. In the end, they would have saved money if they had initially implemented the “more expensive” flash drives.

#3 Footprint — In addition to not needing as many drives to keep performance high, enterprise flash storage devices have a smaller footprint. It’s not uncommon for customers to shrink cabinets into rack units or for SANs to be replaced by virtual arrays with flash being integrated into servers. Flash is also paving the way for Software Defined Storage (SDS) with a smaller and more flexible architecture. For businesses operating a data center in a tight space, this is an important benefit. As an added bonus, a smaller footprint can provide for better air flow, which can also cut down on data center cooling costs.

#4 Energy usage — This argument can be the clincher when the organization has a green initiative or is seeking to lower energy costs. Data centers consume a lot of energy, and according to the Storage Networking Industry Association, 10 – 40% of that energy is consumed by storage devices. Enterprise flash drives can help you reduce costs because they draw less energy. They also produce less heat, further reducing the ongoing cooling costs associated with your data center.

Looking Ahead

Motivated by larger data sets, expectations of faster response time, and shrinking budgets, I believe we have reached the point where flash technology should be the default choice in drive technology. But what I am personally most excited about is what that will mean for our customers. The increase in performance has enabled disaster recovery and business continuity with recovery times in minutes instead of days and ever-shorter recovery points.

[You might also enjoy: Offsite Backups Are Not a Disaster Recovery Plan]

Realizing the promise of Big Data, IoT, and analytics require more than just data collection. Businesses need to be able to access data, analyze it, and serve it up real time in ways that make it meaningful to the audience. Pre-flash, disk access times were commonly measured in milliseconds — often a dozen or more. With flash, sub-millisecond is the new norm. The ability to store and access more data without compromising performance will pave the way for a new generation of applications for business with advanced analytics  or even the next addictively fun game to be installed on your smartphone.

foschini_travis_Travis Foschini, Solutions Architect with TierPoint, thrives in the fast pace of IT  and  desires to be truly consultative in his role. He likes to keep his finger on the pulse of the latest trends around virtualization, storage and cloud services and especially enjoys sinking his teeth into a good availability or performance challenge. Travis focuses his energy on aligning clients’ and prospects’ business and technical requirements with solutions that will make a truly positive impact for their company. 

2 thoughts on “Making the Case for All-Flash Storage

  1. Great article, Travis! I would add that with technologies like inline duplication and compression (enabled with All-Flash Storage), there is even more opportunity to reduce cost and footprint (#2 & #3). This can many times be achieved with little effect on performance (#1).

    Like

    • Note from the Author, Travis Foschini:

      Good point Charles. Compaction technologies enabled by Flash dramatically change the economics.

      The use of Flash in arrays & the addition of more/better effective CPU’s has been used progressively in recent years to improve compaction rates and perform the compaction faster. Asynchronous background processes are inefficient. Compaction can now occur inline & rates of dedup continue to rise to the point of being a key differentiator among products. Storage technology has quickly become about the software layer managing the storage as much as it’s about the disks themselves. The software in turn brings many additional benefits beyond those already mentioned . While the innovation in storage is largely driven by flash, it has so many implications that the reasons are numerous & very compelling for all involved (Buyer, Operator, Application User, etc).

      Like

Comments are closed.