A clear vision of objectives for the present and future of the center should be elucidated at this stage.
Design and building are obvious and crucial aspects that also take into account and depend on the initial analysis and planning phase. One always has to take into account the fact that data centers will often need to expand. This can be catered for by a modular design and building approach; which means that additions for future growth can be built as required.
Relocation is an important part of the life cycle and involves considerations such as the migration of data. This aspect also has to take into account customer or client needs, which falls within the ambit of the managerial function. A data center also has in the course of its life cycle to undergo periodic evaluations. This is necessary to include updates in terms of data increases and also serves as "…an early warning for future changes to the data center that will be required" (Data Center LifeCycle Management. A Comprehensive Solution for the Life of Your Data Center).
3. Data Centers and Their Significance
As has been noted in the above section, increased demand from the public and private sectors and from the growth of business and communications initiatives has placed as great deal of emphasis on the effective and efficient running and management of the contemporary data center. This refers to the exponential growth in the demands and requirements of new it infrastructures which are needed to drive mission critical imperatives and operations. As a number of studies note, there are a wide array of reasons that can be given for the contemporary importance of the data center.
A central reason is that the "… recent business trends such as globalization, governance issues, and heightened security concerns along with technology trends" are placing increasing demands on it infrastructures; at the same time there is a concomitant increase in it expenditure ( MacKinnon, 2008). As a result of this situation there is a growing need for new and more scalable datacenters, which are also more cost effective. As MacKinnon ( 2008) emphasizes, these demands are straining the present design and structure of data centers as well as managerial expertise in running these centers. Consequently,
IT and facilities personnel must work in tandem with the business stakeholders they support to ensure the complex tasks of data center planning, design, relocation, and construction have the flexibility to reflect companies' ongoing and future needs while remaining cost-effective.
Some of the central aspects relating to this view and assessment will be discussed in the following sections.
Data center virtualization refers to "…a method of moving information storage from physical servers to virtual ones, often in a different location" (What Is Data Center Virtualization?). Virtualization differs from the previous more cumbersome methods of data storage with physical servers on the site of the data center holding massive amounts of information. With data center virtualization"…it became possible to separate both the hardware and location from the data. This cuts costs and increases the data's availability"( What Is Data Center Virtualization?).
It should also be noted that there are different degrees and types of data visualization. This may range from web services on a shared server to multi-server data center services with complex applications ( What Is Data Center Virtualization?).
Virtualization has in fact a fairly long history in computing. For example, mainframe computers have in the past used a form of virtualization to partition the machines and run multiple applications (O'Hanlon, 2008). However, it was realized in the 1990s that;
… it was more cost-effective to use virtualization technology to harness the unused power of dedicated servers and run applications across multiple servers than to dedicate an entire server to one application and possibly leave a lot of computing power untapped.
As the amount of data increased so the need for more sophisticated methods and virtualization applications began to emerge, as an integral part of the data center life cycle.
Storage virtualization can take place on two levels; block storage and file storage ( Antonopoulos, 2005). Computing virtualization is another developing area with the use of cross-platform server virtualization software, which provides data center managers with the facility to cluster and partition servers ( Antonopoulos, 2005). Networking virtualization allows network managers to "…create isolation and segmentation into virtual LANs running on a single physical LAN"( Antonopoulos, 2005).
Virtualization also has a great deal to do with customer or client needs and requirements. This is due to a great extent to the significant cost savings that result from this method. The concept and praxis of virtualization has had an impact on the data centre life cycle as well as on management of this process. As one study of a data center for a school district notes; by implementing virtualization they have "….been able to reduce the number of its storage server racks from five to three, which not only decreased the amount of space the district needed for equipment but also cut its energy consumption" (O'Hanlon, 2008). Cost and space saving was also increased in this particular data centre through virtualization. "In addition, the district, which has 22 buildings and roughly 70 data closets, was able to consolidate nearly all of its data storage at its central office using a Xiotech (www.xiotech.com) storage area network (SAN)" (O'Hanlon, 2008).
The above article focuses on the savings in terms of costs to the school district budget. The director of network services states that, "Using server virtualization has saved our school district $7 million in costs within the last year," (O'Hanlon, 2008). Virtualization therefore introduces scalability as well as stability to the containment of data in the life cycle. Another positive aspect that is mentioned in the literature is the "…ability to scale our power consumption based on our system usage"(O'Hanlon, 2008). A further positive factor is that virtualization also tends to reduce the necessity for repairs, therefore making it cheaper in the long run to deal with data recovery.
One of the most important changes with regard to management and the data center life cycle, is that virtualization provides a more scalable and cost-effective means of dealing with disaster recovery. As the manager of a large data center notes;
We found that we don't need to have as large a [plan] for disaster recovery due to the nature of virtualization software… Whereas previously storage was "all over the place," that's been redesigned to fit into two separate storage clusters, one for server data storage and the other for user data storage. That wouldn't have been possible without the server consolidation that could take place through virtualization.
5. Backup and Security
The issue of backup and disaster recovery have already been touched on above. This is an essential part of data security and forms an integral part of the data center life cycle. In essence, "…today's data centers have to support recovery, back-up and redundant systems, as well as remote operations in different geographical locations "(Data Center LifeCycle Management). A backup and disaster recovery process therefore forms a cardinal part of business and data continuity.
The protection of data via firewalls and other methods against any breach of privacy is an obviously essential part of the life cycle and management of data. However, research indicates that security virtualization is a problematic area. "Security virtualization is probably the one area where the technology is still lagging" (Antonopoulos, 2005). As one study emphasizes;
There are a few examples of virtualized security devices that offer multiple instances on a single physical chassis. This can be extremely useful for service providers managing multiple customers. However, we have yet to see broader deployment within enterprise environments.
The future of managing a data center will therefore have to take these factors into consideration as well as the synchronization of virtualized security with other data center applications (Antonopoulos, 2005).
An issue that should be noted is the increasing threat of identity theft and the way that this impacts security during the life cycle of the data center. Prosch (2009) examine methods of prevention identity theft or PI in the data center cycle. The author states that in data life cycle management (DLM)
Although data may lose its value to an organization, in many cases, it does not lose value to identity thieves….Organizations that protect only newer PI with the latest encryption and privacy enhancing technologies are placing their former and current customers at risk. They need to diligently protect all PI, regardless of age and storage medium (Prosch, 2009)
Therefore, identity theft can be prevented in the neither first instance by nor colleting and storing unnecessary personal information. Furthermore, "Once PI gets into the data life cycle pipeline, the cost of managing and destroying it escalates" (Prosch, 2009). Both preventive and detective protection of stored data should be implemented…