Sizing an IT infrastructure today can be tough
Although science is used to determine data center needs, there is still an art that goes into the decision process. The reason is simple: a variety of IT devices of various ages from various vendors — each with its own lifecycle, costs, and on-going maintenance issues — often present unforeseen consequences.
Most IT equipment is expected to actively contribute for at least 3 years without any additional purchases. But it is difficult to anticipate things like IOPS and capacity one year in advance, much less 3 to 5 years ahead. And how do you plan for future technologies that haven’t been invented yet? Faced with so many variables, it’s surprising more IT professionals don’t consult Ouija boards before they make sizing decisions. All joking aside, sizing your infrastructure does include an element of fortune-telling, and some amount of risk is inherent in the process.
Too small, too large, too many changes
If you select a small system with a low upfront costs, you could have problems down the road. Your end users may become angry because they don’t get the performance they expect. Additionally, many under-sized infrastructures require forklift upgrades a year later, eliminating any initial cost savings. And those costs can be compounded. To compensate for the inadequate size of the initial deployment and to guarantee the subsequent upgrade won’t fall short, departments will go to great expense to oversize the environment, well aware that the business won’t be able to take full advantage of the new products for years.
Organizations that look too far into the future will intentionally round up to a solution that is too large for their needs hoping to stay ahead of these sorts of forklift upgrades. This strategy is no better than under-sizing. Imagine the cost of over-investment if the director of IT, VP of IT, and CIO each round up the configuration to mitigate risk. Planned and purchased with the best intentions, these systems can become outdated before the value is ever realized.
Even the most accurate predictions can go wrong when business priorities change. Today’s business models change far faster than the average lifespan of enterprise IT equipment, requiring new levels of agility. Legacy products with a 5-year lifecycle become process and innovation anchors, keeping the business tied to the time period when the technology was acquired.
Don’t be fooled by cost, capacity, or cloud
It’s easy to run the risk of falling into the over/under sizing trap, lured by the lowest upfront costs or the greatest capacity for the dollar. Even public cloud, with its elastic approach to infrastructure, can be expensive in the long run. Continuously adaptable to meet business needs, this flexibility comes with trade-offs, including cost, compliance complexity, performance based on user and data locality, and data sovereignty. Although you can grow your environment in very small increments and only when needed, costs often go up very quickly if sizing is off.
Don’t be drawn into a quick decision based on a system that promises to deliver top results five years from now. Instead, reign your sights in to the next 1-2 years and assess your current state before making any decisions. Do you need all-flash performance to support tier-1 apps? Do you have remote and branch offices with sizing requirements of their own? Take a good look at your business needs and size your infrastructure according to your current workload and use cases, then find a solution that leaves options open for future growth.
Sizing for your environment
Each environment has unique data capacity and performance requirements, which is why one size won’t fit all, even in the most flexible, agile infrastructure. VDI is a great example of where sizing can be very difficult to anticipate and potentially costly to address, if the core system proves to be inadequate. The performance needs of the desktops in a new VDI environment, even when science is applied, cannot be accurately anticipated. As ESG points out, predicting what these performance needs will be, how quickly users will adapt, or which user types will be best suited is challenging, especially when planning a VDI environment.
Hyperconvergence offers a cost-effective, scale-as-you-need architecture that is particularly well-suited to VDI and remote office deployments. While it will not provide you with the ability to predict business needs 5 years in advance, it will allow you to set up a reasonably sized infrastructure for the next year and grow in a direction that makes sense. Every hyperconverged node includes all the components needed, so cost and capacity are easy to calculate. These smaller building blocks don’t require large upfront investment or forklift upgrades, so customers can adapt easily to changing dynamics.
Better sizing with hyperconvergence
Sizing is still important, because you want the overall investment to last 5 years. Hyperconvergence offers scalability that is well-defined and simple to implement, allowing customers to start with a simple proof of concept to better understand the viability of a technology and the performance envelope it delivers, then grow the environment as needed.
No crystal ball can ensure perfect sizing, but with hyperconvergence the costs are well-known up front so contingency budgets can be planned for. And hyperconvergence allows the performance to grow linearly, making capacity planning simple. HPE SimpliVity hyperconverged solutions offer a huge advantage for businesses that want the agility and incremental cost advantages of the public cloud in their own data center.
Examine the total economic impact of deploying a hyperconverged solution in this Forrester report.
About Jesse St. Laurent
Jesse St. Laurent is the Chief Technologist for HPE Hyperconverged and SimpliVity. He uses his 20 years of experience to engage channel partners, evaluate emerging technologies, and shape innovative technology solutions involving data center modernization. For more information on how hyperconverged infrastructure can elevate your hybrid IT environment, download the free HPE SimpliVity edition of Hyperconverged Infrastructure for Dummies ebook.
To read more articles from Jesse St. Laurent, check out the HPE Converged Data Center Infrastructure blog.