Six Barriers to the All Flash Data Center
By Doug Ko – Strategy and Market Development
With the release of the all new Nimble Predictive All Flash arrays, it’s now easier than ever for organizations to implement cost-effective all-flash data centers. To start, the new Nimble arrays bring down the cost of flash by 33% to 66% compared with other all flash products.
But experienced IT leaders know there are often many important considerations other than price when implementing new technology, and that’s definitely true of flash storage. In particular, our research has identified a number of potential roadblocks that can slow down and hinder your transition to flash storage.
Here are six major barriers to watch out for:
Application performance bottlenecks – With flash arrays, high performance is a given – that’s what flash is for. But what about identifying and avoiding the slowdowns that occur in parts of the infrastructure stack other than storage? Our data show that 54% of these slowdowns are not caused by the storage system, but are issues with hosts, networking, interoperability, and failure to follow best practices. Nimble’s Predictive Flash platform goes beyond storage monitoring, and helps you understand performance across the entire infrastructure stack.
Poor availability – Every IT leader wants to deliver no downtime, no data loss, no breaches. And it’s easy for vendors to promise something like “built for 99.9999” theoretical claims. Look for real data, with measured uptime from production arrays. Nimble currently clocks in at 99.9997% measured uptime across our more than 7,500 customers, and continues to improve.
Managing storage silos – You’re not going to flick a switch tomorrow that turns your entire data center to all flash. Instead you need to manage different storage technologies through the transition to all flash. All flash suits latency sensitive applications, but what about those other mainstream applications that need something more cost-effective? You need a storage platform that can seamlessly manage and move data between hybrid disk-flash and all flash – a single consolidation architecture spanning all flash and hybrid flash arrays.
Limited scalability – Not Nimble. We scale to 1.2 million IOPS and 8 petabytes of all flash. But it’s not just about big numbers, it’s about flexibility of scaling. Look for storage that allows you to independently scale-up the performance and capacity of an array and to scale-out by clustering multiple arrays, achieving linear performance and capacity scaling.
Disaster recovery costs and risk – Flash is great, but it’s still expensive to use it for backup and disaster recovery. Sure you can use cheap disk or tape and software to backup all flash arrays, but this adds cost, complexity and risk to your recovery plan. Keep it simple and save costs by looking for a platform that replicates from all flash to cost-optimized hybrid flash without needing third party software.
Forklifts, hidden costs, and the cloud – You’re looking to invest in all flash. Don’t let that money go to waste on products that nickel and dime you on extra software features, or strong-arm you down the road into forklift upgrades by raising maintenance costs. And with the cloud looming above, you can’t always afford straight capex purchases or leases, and need to look for pay-as-you-go alternatives for flash storage.
For many organizations, the question is not if they’ll move to an all-flash data center, but when. Planning around the barriers above will put you on the right path and get you there faster. For a detailed guide on how to cost-effectively transition to flash storage, download the free eBook, Overcoming the Obstacles to the All Flash Data Center.
- Doug Ko