The acceleration of business change is putting organizations under pressure to deploy the best-performing and most cost-effective IT services available. After all, information is the most highly-prized asset of any organization, and as they accumulate greater amounts, things become much more difficult to maintain.
Employees are accessing more data, and in doing so are using more devices than ever before. This calls for new solutions as organizations strive to keep pace and stay within their budgets. Successful organizations realize the vital importance of being able to store, secure, access and analyze their data on demand, and this demand is driving a number of trends within enterprise storage that are set to heat up over the months and years ahead.
Flash storage economics
One of the hottest trends right now is Flash, with IDC predicting it will account for 17 percent of spending in enterprise storage in the next three years. That’s because Flash is up to 1000 times faster than traditional hard disk drives, while offering lower latency and higher throughput.
Flash won’t completely rule the roost just yet though, thanks to its prohibitive cost. While it can be deployed in the servers or in storage arrays, doing so requires faster server-to-server or server-to-storage connections. And while Flash is cheaper on a cost/performance basis than HDDs are, on a cost/capacity basis it’s still five-to-ten times more expensive. Many organizations are therefore only deploying Flash in their most critical areas, rather than “wasting it” on slower networks.
The rise of Server SAN
The second-hottest trend right now is Server SAN, which involves sharing and distributing storage across multiple nodes. Wikibon defines Server SAN as an “architecture that turns multiple direct-attached storage (DAS) devices into a pool of shared resources via a high-speed interconnection such as InfiniBand or Low-latency Ethernet”. Although Server SAN is in the very early states of penetrating the enterprise storage market, Wikibon says it’s poised to disrupt it completely, having already succeeded in disrupting the hyperscale service providers marketplace.
Essentially, Server SAN speeds things up for fast and low-latency networks by allowing each node to access remote storage as quickly as if it were local to that node. Examples of Server SAN solutions include EMC ScaleIO, VMware’s VSAN, and clustered file systems such as Lustre, IBM GPFS, Red Hat Gluster, and Quantum StorNext.
Big Data & Cloud
Big Data and Cloud storage solutions share many similarities, being able to support massive amounts of data and requiring numerous servers. That’s why customers, cost-conscious as they are, usually opt for open-source solutions like OpenStack and Hadoop. There are some differences though – Big Data normally uses local storage for each server, although storage can be centralized in order to improve efficiency. Meanwhile, Cloud normally uses centralized storage, but can sometimes use distributed local storage for Web 2.0 deployments and large public clouds.
Both Big Data and cloud will often leverage Flash for higher performance, and this requires fast networks to ingest, share and analyze the data.