Flash is not Wine

Flash is the go-to performance option when IT professionals face a storage performance challenge. The problem is that the price of flash is still at a premium versus what you can buy traditional hard disk technology for. And the reality is that not all (actually most) workloads don’t need flash. As a result organizations have to make a choice: do they overbuy on flash, an all-flash array, or try to manually balance cost and performance with a hybrid storage system?

The All-Flash Problem

All-flash arrays are the performance sledgehammer. They don’t solve performance, they obliterate it. Again, the reality is that most workloads don’t need flash performance, at least not all the time. Flash storage you use for something other than solving performance problems is a waste of the technology’s potential and of the organization’s money.

Ideally an organization should be able to buy just the amount of flash they need right now and incrementally add more flash capacity as needs justify it. Most all-flash vendors don’t have an answer to this problem. They typically sell you way more flash capacity than you need, counting on you to fill it up eventually. The problem for the buyer is that flash dramatically goes down in price every six months.

Flash is Technology Not Wine

Flash is not wine, it does not become more expensive with age, it gets cheaper. Buying excess flash capacity NOW is one of the worst investments that an organization can make. But buying the bare minimum amount of flash is also problematic. Most organizations will continue to need more flash as time goes on and workloads scale. Vendors make adding flash capacity to their systems, either through scale-up upgrades or scale-out architectures, very easy. The problem is the human element as the upgrades occur. Since you have to budget and pay for additional capacity, some would say it is justifiable to do a massive upgrade on a schedule and use the available IT staff to make sure the upgrade or expansion works. The theory is, once you add capacity later, you may need to reconfigure workloads so users can access the new capacity. But there is a way around this so you can get the flash you need now, affordably get it later and not have to reconfigure everything to make sure you can access that new additional capacity.

Introducing Flash-As-A-Service

You can see in an on-demand webinar we did with Tegile that you can address these challenges in a new way to solve the performance versus budget problem that flash causes. It is essentially a new business model. Tegile will install the flash storage system on-premises with more than enough capacity to meet the organization’s needs but the organization only pays for the exact capacity that they use, not more. If the organization’s needs grow, it can quickly unlock the existing capacity without having to schedule an upgrade. More importantly if their needs decrease they can give capacity back to the vendor. Not even scale-out storage systems can do that.

To learn more watch our on-demand webinar “Flash as a Service” and see if this new way of acquiring flash makes sense for you.

Watch On Demand

Watch On Demand

Twelve years ago George Crump founded Storage Switzerland with one simple goal; to educate IT professionals about all aspects of data center storage. He is the primary contributor to Storage Switzerland and is a heavily sought after public speaker. With over 25 years of experience designing storage solutions for data centers across the US, he has seen the birth of such technologies as RAID, NAS and SAN, Virtualization, Cloud and Enterprise Flash. Prior to founding Storage Switzerland he was CTO at one of the nation's largest storage integrators where he was in charge of technology testing, integration and product selection.

Tagged with: , , , , , , ,
Posted in Blog
One comment on “Flash is not Wine
  1. WaltR says:

    IMHO, if there every was a time to think differently about All Flash purchases, it is now.

    Non-volatile storage is facing it’s biggest technology disruption since MLC appeared, maybe since NAND appeared. Almost any All-Flash system you buy today will be built with last year’s flash technology. Especially if they are using anything other than off the shelf hardware.

    With other NAND manufacturer’s getting through their teething problems with 3D Flash and starting to compete with Samsung in that market, price/performance should accelerate faster this year than last.

    3D xpoint and other emerging non-volatile memory technologies will be challenging Flash on the high end. Next year’s buy will be faster and cheaper.

    Unfortunately, by 2018, you may not be able to expand the product you buy today. I expect, pure-play all flash vendors could be looking to be bought or restructuring in 2018 when their secret sauce looses it’s value. Vendors that continue to rely on high dedupe rates (and it’s inherent CPU overhead) to meet “effective storage” requirements will be the first to feel the pain as flash cost/TB accelerates downward and reliability increase unless they pass the savings to customers. If you can buy a system with raw capacity you require at same price/performance before implementing compression or dedupe, why take a chance on a system that may or may not have the capacity you need depending on your use case?

Comments are closed.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 22,231 other followers

Blog Stats
%d bloggers like this: