Data centers come in a variety of shapes and sizes which can be located onsite or in a large facility offsite. The data center for a small business or school may only occupy a room that is a 500 square feet, while some of the largest data centers may occupy an entire building over 1 million square feet. Selecting the appropriate size data center will depend upon many factors including current and future requirements, budget, uptime requirements, and location. Making the appropriate selections while the data center is being designed can help avoid costly reconfigurations or outages in the future.
It is critical when selecting a data center size that it meets the current requirements of the organization. However, it is also important to consider the future growth of the organization. It is important that the room or building housing the data center has additional space available for equipment that may be needed in the future. It is also important that the power, HVAC and fire suppression systems can be upgraded. As the equipment in the data center expands, these supporting systems must also grow to a size that will support the new equipment.
When considering all of the components of a data center, the cost can quickly surpass the initial budget. There are many different ways that costs can be controlled. Many manufactures offer equipment that accepts modules, such as network switches that accept SFP modules. This allows the network administrator to custom tailor the equipment to the current needs of the organization, while providing a path for future upgrades and expansions. Another area where cost savings may exist is the amount of redundancy that is built into both the power and communications system. The network administrator must consider the uptime requirements of the data center when selecting which type and how much redundancy will be built into the system.
There are four major tiers of data center uptime requirements; basic, redundant, concurrently managed and fault tolerant. Basic or Tier 1 data centers provide for 99.67% availability. Data centers built to this level will have a single path design for power with non-redundant cooling and components. These data centers may experience up to 28 hours of downtime per year. Redundant or Tier 2 data centers provide for 99.741% availability. Data centers built to this level will have a single path for power with non-redundant cooling and redundant components. These data centers may experience up to 22 hours of downtime per year. Concurrently Managed or Tier 3 data centers provide for 99.982% availability. Data centers built to this level will have dual paths for power, cooling plants and redundant components. Generally, only one path is active at any given time. These data centers may experience up to 1.5 hours of downtime per year. Fault Tolerant or Tier 4 data centers provide for 99.995% availability. Data centers built to this level will have fully redundant equipment, components and signal paths. Multiple paths will be active at any given time ensuring that a single failure will not cause an outage. These data centers may experience up to 26 minutes of downtime per year.
It is important to select the right location for a data center to help avoid a loss. The location for a data center should include plans to avoid natural disasters as well as potential accidents. To avoid natural disasters, a data center should be located in an earthquake Zone 3 or less, away from flood plains or dams, where hurricanes and tornados are infrequent. To avoid potential accidents, a data center should be located away from major highways and railway lines, and away from hazardous production facilities, airports and flight corridors. Placing the data center in the proper location will help prevent potential catastrophic events.
This white paper is for informational purposes only and is subject to change without notice. C2G makes no guarantees, either expressed or implied, concerning the accuracy, completeness or reliability of the information found in this document.
