![]() ![]() English
|
3PAR® Adaptive Optimization software gives enterprise and cloud datacenters the ability to optimize service levels autonomically, on a large scale, and for a lower total cost than any other solution available today.
Adaptive Optimization takes a fine-grained approach to autonomic storage tiering that optimizes service levels by pairing data at the sub-volume level with the most cost-efficient resource capable of meeting its particular service level requirement. Policy-driven, granular data movement takes place autonomically, on an ongoing basis—so the right Quality of Service (QoS) is delivered to the right data at the right time, at all times.
Sub-volume autonomic storage tiering with 3PAR Adaptive Optimization software enables enterprise and cloud datacenters to flexibly meet even the most demanding and dynamic service level targets for up to 30% less than using Fibre Channel drives alone. By analyzing and moving data to the appropriate storage tier at the sub-volume level, the storage array can use an extremely “lean” premium SSD tier to service only the most performance-intensive data at any given point in time. The InServ’s abundantly scalable levels of performance and ability to stripe writes widely across all array resources enable the use of highly economical, widely striped Nearline (enterprise SATA) drives to meet broader capacity requirements—achieving service level targets at the lowest possible cost.
3PAR Adaptive Optimization leverages the same proven data movement engine used by 3PAR customers for online, non-disruptive rebalancing of storage volumes across 3PAR InServ® Storage Servers with 3PAR Dynamic Optimization software. In addition, Adaptive Optimization incorporates several key policy override mechanisms to protect against user impact. IT managers maintain complete control through a flexible implementation that allows data movement to be scheduled as well as resource usage limits and tier definitions to be varied by application.
3PAR Adaptive Optimization software also includes QoS gradients that can be used to bias data movement within a profile based on specific performance or cost objectives. A QoS gradient accelerates or decelerates data movement toward a particular class of resources so the user can better meet service level and cost objectives. For example, a performance gradient can be used for data with high service level demands, such as a seasonal order management application. As workloads begin to spike (such as when Black Friday or Cyber Monday approaches), the gradient will rapidly and autonomically move data to high performance resources and maintain it there until after the activity has declined.