A closer look at capital costs: Traditional vs. scale-out storage in a data center

By Igal Smith

Over the past several years, scale-out storage technologies have grown increasingly popular. Many companies have considered the technological advantages scale-out storage provides in a data center; however there have been few economic comparisons. This makes it difficult for many companies to decide whether or not the potential benefits of scale-out storage in a data center will translate into cost savings. The first question every business asks when investing in new technologies for their data center is whether or not the initial capital costs allow for it to even be considered. To compare the overall capital costs of traditional and scale-out storage in a data center, several variables must be considered.

Pricing

Traditional storage in data centers often rely on proprietary hardware components. Scale-out storage environments can be built using off-the-shelf servers. Commodity pricing is always lower than proprietary pricing because market pressures constantly drive down commodity prices whereas vendors can always charge a premium for proprietary storage systems in data centers. On a per unit basis, scale-out storage in a data center is much less expensive over the lifetime of a business.

Total investment

 

Traditional storage systems in data centers allow users to scale up, however they must invest in expensive hardware initially deemed appropriate for the eventual level of scale. This inherently creates a significant upfront investment because of expected future storage needs.  Plus, it encourages over provisioning which does not provide adequate returns if it takes longer than expected to reach the estimated capacity target.  In a data center, scale-out storage allows businesses to add new nodes whenever the amount of storage needs to be increased. This allows businesses to start small and grow over time whenever their needs dictate.

Hardware repurposing

A significant financial benefit of scale-out storage in a data center is the ability to repurpose existing hardware.  This defers new capital expenses because decommissioned servers can be added to new and existing storage environments throughout a data center.

High availability

 High availability can be achieved more cost effectively in a scale-out storage environment.  This is because the software ensures high availability and the failure of an individual server will not compromise access to the data or affect the system's overall availability.  This also allows less extra capacity to be maintained.  Traditional NAS storage in a data center relies on a storage chassis which is replicated for critical applications and held in reserve in the event of a hardware failure.  This requires significantly more resources, both in terms of hardware and financial investment.

Potential for additional capital cost savings

Traditional NAS storage in a data center eliminates the possibility of additional capital cost savings because users are limited to a select number of alternatives provided by the vendor.  Scale-out storage in a data center allows users to deploy the exact hardware and software they need. The additional flexibility provides the opportunity for additional capital cost savings on a case-by-case basis. When looking solely at capital costs, scale-out storage is clearly more economically advantageous. Not only does it allow companies to spend less per unit, but it also allows for a smaller initial investment.  Factoring in the ability for hardware repurpose seen and the potential for additional savings, scale-out storage provides long-term financial benefits as well.

Attached Images: 

•  License: Image author owned 
•  License: Image author owned 
•  License: Image author owned 

Igal Smith writes articles about colocation and data centers. He believes that using a colocation service is a wise choice for many businesses. Igal believes companies should always think about customizable solutions that are tailored to meet their needs. The Colocation service he recommends can be found at http://www.datafoundry.com/colocation

Differences between data warehousing and business intelligence

The primary differences between data warehousing and business intelligence are in function and design. Data warehousing refers to the process, mechanism and infrastructure utilized in the preservation of digital information. Business intelligence on the other hand, is the use and analysis of information, some of which may be obtained via data warehousing. This information is used for the decision making pertaining to the operation of a business.

The State of North Dakota describes business intelligence thoroughly in its 'Business Intelligence Report'. Specifically, the white paper describes business intelligence in terms of technology and infrastructure such as data mining information held within a data warehouse and using the appropriate software to perform data analytics. This helps in designing business models that have basis in logical, mathematical and qualitative assessment of business data.

Data warehousing can be thought of as a component of business intelligence that has wider scope. However, without adequate data warehousing, effective business intelligence may have little to draw valid conclusions from and therefore lead to faulty decision making. Thus, data warehousing ideally possesses qualities such as strong organization, easy access, time sensitivity and relevance. Since data warehousing takes physical space used by hardware storage,  practically making use of space, and equipment to store sufficient data also involves prudent determination.

The objectives of data warehousing and business intelligence further delineate differences between the two. For example, business intelligence attempts to define, assess, observe, forecast and calculate for the benefit of a business; this is usually or ultimately financial in nature by design and function of a business. The corresponding function of data warehousing is to serve the interest of business intelligence and other business data storage needs.

Further differences between data warehousing and business intelligence include variation in software applications, employee tasks, cost and hardware needs. For example, an information technology (IT) professional familiar with magnetic, optical and solid  storage devices would more likely be involved with data warehousing than an executive marketing manager seeking to make use of marketing research to implement a new product launch.

The service industry itself is differentiated in such a way as to distinguish between data warehousing and business intelligence. This is because managerial science and information technology become specialized and require specific knowledge in data warehousing and business intelligence.

Differences arising out of, and between data warehousing and business intelligence may also have the potential to lead to professional entry barriers or differences in approach to a business operation.  To illustrate, an IT professional may install and program a data warehouse to store information with maximum capacity in mind, whereas a professional focusing on business intelligence may emphasize strategic use of information instead.

Sources:

1. http://bit.ly/eABq8x (North Dakota: Business Intelligence)
2. http://bit.ly/2ryqwS  (Internet Journal)
3. http://bit.ly/gtLrDV  (National Security Agency)