Balanced performance among IT infrastructure components is essential to the cost-efficient operation of applications. Servers, networks and storage components must work together for optimal application performance. An imbalance may lead to increased costs or reduced performance.
Storage is an important component of many workloads, especially those that require a high volume of input/output operations per second (IOPS), such as server virtualization and virtual desktop infrastructure (VDI).
Removing storage bottlenecks is critical for the performance of these applications, especially as server and networking technologies become more powerful. Storage vendors are developing exciting next-generation technologies such as all-flash storage arrays and solutions that employ Non-Volatile Memory Express (NVMe) to boost storage performance and bring architectural balance to the data center.
As state and local government agencies consider all-flash storage systems (often referred to as solid-state storage), they should keep these tips in mind to achieve optimal results.
1. Consider How Flash Affects the Broader Environment
When planning a storage deployment, a comprehensive approach to determining return on investment (ROI) will deliver better operational outcomes. All-flash storage can be expensive, but the expense should be viewed within the context of the capabilities and value it delivers.
IT leaders should understand that reducing costs for a storage system may result in increased costs elsewhere. For example, storage can directly affect the performance of a VDI deployment. A broad interpretation of ROI would also take into account the productivity of all system users. While a storage system that delivers slower performance may reduce the cost of the VDI deployment, it may also hamper productivity and thereby increase personnel costs.
2. Use Management Software to Address Complexity
Some all-flash storage arrays require specialized software to be installed on all servers that access the storage. While this approach can improve performance and add other capabilities to the storage, it also adds complexity that IT staff must manage. The cost and personnel demands of this requirement should be part of any planning efforts.
Other arrays may use a tiering scheme that places flash storage in an environment along with traditional spinning disk storage. While this can be a powerful approach to expanding capacity, the increased complexity may require additional expert personnel.
Management software that consolidates storage monitoring and control into a single dashboard can help IT personnel deal with increased complexity. Effective management software enables easy configuration of the entire system, includes robust performance monitoring and predictive alerting capabilities, and helps IT staff implement planned system upgrades.
3. Match Storage Performance to Application Requirements
Implementing an architecturally balanced storage system requires a deep understanding of application performance requirements. If an application demands high availability, IT staff should choose a storage system with a rich set of replication, reliability, availability and serviceability features. If an application demands extreme capacity or low costs, a tiered architecture may be a better fit.
NVMe technologies work well in deployments that emphasize IOPS. Drives that employ this technology generally deliver a tremendous number of IOPS — perhaps 400,000 or more per device, compared to hundreds of IOPS per device from a traditional hard disk drive. With such a disparity in performance, it doesn't take long to justify the additional cost of NVMe.