Organizations have historically viewed data storage as a place to stash information away in case it’s needed someday. That approach has led to vast repositories of data that is rarely, if ever, accessed. Industry analysts have estimated that less than 40 percent of stored data in the enterprise is used to generate insight.
It makes little sense to continue warehousing rapidly growing data volumes at ever-increasing cost. Smart organizations are looking for ways to process and analyze data in real time in order to improve decision-making, reduce risk and create competitive advantages. This had led to increased adoption of so-called “storage server” solutions, which provide the performance needed for data-intensive applications by moving storage closer to compute resources.
Introducing the Cisco UCS S-Series
Cisco is facilitating this trend with the introduction of the Unified Computing System (UCS) S-Series storage-optimized server. The UCS S3260 Storage Server, the first in the S-series product line, provides 600TB of local data that can be scaled up to petabytes with UCS Manager. It features a dual-node, two-socket architecture with Intel Xeon processors, and offers cache acceleration and unified I/O connectivity for any type of storage. Capacity can be scaled out to 86PB by adding nodes in a UCS domain.
To create the technology behind the UCS platform, Cisco leveraged its networking expertise to tightly integrate network and virtualization layers. With the UCS S-Series, Cisco has extended that concept to storage, creating a system that addresses the requirements of big data analytics and supports the deployment of software-defined storage, object storage and data protection solutions.
Why Cisco UCS Beats Traditional Servers
The storage server concept has gained traction in recent years as more organizations seek to adopt applications such as video analytics, security analytics and machine learning. Combining storage and compute resources not only improves performance but simplifies management and enables a high level of automation. Storage servers can be built from commodity hardware using direct-attached storage that is connected via a high-speed interconnect and managed through software.
However, not all commodity hardware is created equal. With its modular architecture and the automation of UCS Manager, the UCS S3260 reduces capex by up to 34 percent compared to traditional servers, and lowers ongoing management costs by up to 80 percent. It also takes up to 60 percent less space, consumes up to 59 percent less power and reduces cabling by up to 70 percent. Organizations can take advantage of familiar Cisco UCS Management tools and gain the peace of mind provided by Cisco’s worldwide support and service capabilities.
In addition, the S3260 reduces total cost of ownership (TCO) by more than 50 percent compared to public cloud storage, while enabling organizations to right-size their infrastructure to efficiently power every workload. According to Cisco, organizations that move 420TB of public cloud storage to the S3260 can break even in just 13 months.
Many organizations are looking to move beyond simply warehousing data to leveraging it for big data analytics and other dynamic applications. The Cisco UCS S3260 Storage Server supports these initiatives by delivering the scalability and processing power to analyze and act on data in real time.