“Agility” has become the marching order of today’s digital transformation initiatives across industries ranging from IT to services to fast food restaurants. A vector of “scalability” and “adaptability,” the term simply means: “the ability to respond quickly to changes or threats.”
An agile data center is therefore one which can scale or downsize quickly with growth, respond well to variable workloads and get back on its feet quickly after a network outage or system failure. This comprises a few distinct goals:
- Meeting service level agreements (SLAs) by maintaining performance standards
- Reducing the amount of labor and steps required to service, replace or scale systems
- Reducing the impact of failure on operations using automation and straightforward response strategies
If this scheme smacks of getting your cake and eating it too, that’s because efficiency and higher performance rarely go together with a reduction in labor or complexity. DCAs often walk a thin line between cost and performance, falling back on redundant infrastructure and a large staff to keep things working smoothly.
Why Data Centers Aren’t Agile
The average data center has too many moving parts to be predictable. Equipment failure happens all the time, whether due to mechanical issues or human error. When it occurs, deployment is often cumbersome, resulting in a lengthy time-to-service (TTS).
To make matters worse, most data centers are perilously complex, involving an indefinite number of blades, racks, switches and routers without a central administration platform. This “siloing” of equipment and systems means that upscaling or downsizing is no easy task for administrators.
Agility means – among other things – efficiency and simplicity. When an organization, strategy or workflow devolves into a rubik’s cube of interconnected parts, one small change has unintended consequences, creating resistance to change in general.
Fortunately, it is possible to achieve even the highest ambitions of agility in the data center using a non-traditional approach to IT: unified computing.
The Unified Computing Servers
Cisco’s unified computing system (UCS) is a converged IT architecture for enterprise environments which abstracts hardware from functionality, allowing administrators to define and automate processes with minimal labor.
Using fabric interconnects (FI), UCS takes a “single pane of glass” approach to server management, enabling administrators to manage equipment through the creation of policies, pools, and profiles for blades and rackmount servers. Not only does this centralized approach eliminate a significant amount of grunt work, it also solves the issue of siloing in fell swoop.
Here are some of the ways that UCS creates a more agile data center:
- Policy Based Management allows administrators to configure 100 servers as easily as configuring one. UCS allows for the creation of service profile templates which are automatically configured as soon as a device is plugged in. This approach is as close to “plug and play” as a data center has ever come, leading a significant reduction in labor.
- By eliminating the need to manually boot and configure servers, UCS reduces TTS by a significant margin. In comparison with competitors, blades can be fully integrated into UCS fiber interconnects 47-77% faster than in traditional server environments.
- By simplifying the number of switching layers, devices and equipment connected to a network, UCS also allows for fast scaling and easy downsizing. Adding and removing devices is painless, requiring no manual adjustments.
- UCS improves utilization of equipment through the use of server pools that dynamically repurpose to accept certain kinds of traffic or tasks depending on workload. This reduces the amount of equipment a data center must deploy, leading to lower costs.
While the term “agile” has been overused in recent years, UCS is one of the few technologies which really lives up to that promise in a field where agility has never been more elusive or important. And with Cisco’s recent updates to its UCS server line configured to support edge computing and AI/ML workflows, support isn’t going anywhere soon.
If you’re looking to bring digital transformation and agile processes to your data center, UCS isn’t a bad place to start the search.
Professional Cisco Support in Northern Virginia
Located near the booming data center corridor in Loudon County, VA, Digital Tech Inc provides rapid response maintenance services including EOSL extension, spare parts, short- and long-term maintenance agreements, migration assistance and depot repair options. Our skilled engineers offer multi-vendor support, covering IBM, HP, Dell EMC, Cisco, NetApp, and many more.
To learn more, contact us today.