Today’s scale-up storage architectures are inflexible, requiring upfront forecasting of performance needs over an array’s three-to-five-year lifespan and creating separate storage silos that complicate management. Scale-out cluster solutions provide upfront flexibility, but tie performance and capacity together, requiring customers to incur higher incremental costs every time they add a storage node.
Nimble Storage addresses these challenges by allowing customers to purchase exactly what they need up front and by providing the industry’s most flexible scaling options. Customers can thus protect their existing investments while growing their storage at the lowest incremental cost without downtime as their needs evolve. Nimble Storage offers three flexible paths for scaling:
(1) (1) Scale capacity only: a new line of ES-Series storage expansion shelves allows customers who do not require additional performance to add capacity at the lowest incremental cost without downtime.
(2) (2) Scale performance only: a new line of extreme-performance CS400 series arrays is ideal for customers running performance-intensive applications such as OLTP and VDI. Existing CS200 series array controllers can be nondisruptively upgraded to CS400 arrays without downtime. Customers can also upgrade to higher-capacity flash SSDs without downtime to accommodate workloads with larger active data sets.
(3) (3) Scale capacity and performance: a powerful new operating system upgrade, Nimble OS 2.0, allows customers to cluster arrays together, providing linear scaling of both capacity and performance, as well as unified management. Nimble scale-out clustering allows users to grow or shrink their storage environments seamlessly, and to easily perform data migrations and upgrades, all without downtime.
Enterprises can mix and match these approaches to scale to hundreds of terabytes and hundreds of thousands of IOPS in a single storage cluster.
“Traditional scale-up systems and more modern scale-out systems are rooted in an era when storage capacity and storage performance were tethered together,” said Suresh Vasudevan, CEO of Nimble Storage. “Our scale-to-fit architecture delivers an unparalleled ability to independently scale the controller performance, cache capacity or storage capacity of any node while also allowing multiple nodes to become part of a cluster. Starting with a small footprint, our customers can continually and nondisruptively scale and evolve their infrastructure in small, granular increments across the widest range of workloads.”
Nimble’s scale-to-fit technology is built on the groundbreaking Cache Accelerated Sequential Layout Architecture (CASL). CASL is architected from the ground up to leverage flash and high-capacity disk to deliver affordable performance and capacity. In addition, CASL delivers highly efficient snapshots and WAN-efficient replication, dramatically simplifying backups and disaster recovery. Operation of the arrays is simplified across their lifecycle through deep application integration as well as proactive wellness and management.
“Enterprise storage requirements are continually and rapidly in flux,” said Roger Cox, research vice president at Gartner. “Consolidation of diverse workloads means that higher demands are placed on storage. Because most enterprises can’t predict their storage requirements even one year down the road, they need assurance of elastic performance and capacity in their storage platforms.”
Powerful scale-out clustering with Nimble Storage allows enterprises to size their storage to current requirements, eliminate the need to forecast future requirements and eliminate storage silos:
• • Scale performance and capacity linearly with the seamless addition of new arrays
• • Automatically “stripe” storage volumes across multiple arrays in a cluster
• • Create storage pools in a cluster to segment applications and workloads on different arrays
• • Migrate data nondisruptively across pools in a cluster
• • Add arrays to a cluster or remove them with full, uninterrupted data access. The Nimble operating system automatically redistributes data across arrays in the cluster.
Furthermore, scale out is compatible with all existing and new Nimble Storage arrays, and any arrays can be mixed and matched within a scale-out cluster. Nimble makes scale out available at no cost to all customers with current service and support contracts.
“Nimble exceeded our most optimistic expectations for data throughput in a pair of mirrored units that were easy to install, configure and deploy,” said Paul Dryden, IT operations Manager at Morgan Cole. “Nimble is an ideal solution for supporting legacy databases, deploying VDI and providing instant recovery from failure for our servers.”
CS400 Series Extreme-Performance Arrays
Nimble’s new extreme-performance CS400 arrays are designed to host an enterprise’s most demanding applications and workloads, supporting hundreds of virtual machines or thousands of VDI users on each array.
“Nimble’s CS400 provides a single storage server that can meet the requirements of virtualisation on a large scale, allowing enterprises to support their core applications, workloads and users, while also providing room to grow,” said Raj Mallempati, director of product marketing, VMware. “End users using VMware View™ solutions are looking for uncompromising performance at scale. Nimble CS400 and VMware View provide a high-performance, scalable desktop-virtualisation solution that can meet the ever-growing capacity requirements of enterprise users.”