Panelboards, Right-Sizing Data Center Redundancy as a Means of Energy Management
Data centers generally act as large warehouses, which store a company’s IT equipment, and important data. As such, these installations must be highly protected from adverse events such as ‘brownouts’ which may cause component damage, or result in data loss. A method which is utilized by most facilities to protect equipment and panelboards and saved data is redundancy.
Redundancy basically refers to non-utilized data center components which are connected to the existing system, and implemented only as backups in the event of equipment failure. Redundant components may include servers, hard drives, coolers and power systems, as well as other equipment depending upon the centers needs.
Many companies pride themselves on their system design, and as such data centers have been divided into four different ‘Tiers’, which basically revolve around the amount of ‘uptime’ that the facility is able to achieve. This, in turn, can be directly linked to the amount of redundancy within the system. Redundancy within the data center tier system is consequently defined as:
- Tier 1 – basically no redundant components (generally can’t sustain more than a 10 minute power outage without down time).
- Tier 2 – some redundancy in power and cooling systems (able to sustain a 24 hour power outage without downtime).
- Tier 3 – redundant power, cooling and service providers (able to sustain a 72 hour power outage without downtime)
- Tier 4 – full system redundancy (able to sustain a 96 hour power outage without downtime)
Issues with Redundancy in Data Centers
Many companies look at these figures and assume that ‘Tier 4′ facilities are preferable, owing to the fact that they then don’t have to worry about component damage, data loss and expensive downtime in the case of an adverse event. This however, is not the case for many businesses due to the massive expense associated with such a high level of redundancy within a data center facility.
Redundant Even with panelboards
The expense generated due to redundancy generally stems from the fact that centers must keep excess redundant equipment such as power supplies, panelboards and servers powered up during normal operations, which due to the rising cost of energy, is basically unaffordable for smaller operations. Centers must also consider the impact that high energy utilization has on the environment, owing to the current implementation of carbon taxes and other associated regulation.
Right-Sizing the Level of Redundancy within a Data Center
The energy efficient and environmentally sustainable solution for many installations consequently lays in right-sizing the level of redundancy within their facilities, in order to gain the required uptime with minimal expense. Smaller scientific facilities for example do not require a high level of redundancy, as downtime is not a critical event for users. Large cloud computing facilities in comparison must provide users with dependable, constant services, which in turn, calls for a higher level of redundancy.
In conclusion, it can be noted that the reduction or consolidation of redundant components within data centers is an easy way for companies to reduce their energy utilization. It is believed that the rising cost of energy no longer allows for facilities to operate with higher than required levels of redundancy, which therefore requires businesses to properly right-size redundancy within their IT installations.
