Electrical power is a critical element in every reliable data center/server room and directly affects the level of system availability. Several key power components exist in a data center/server room design. These include the main electricity power source from the utility company (CLPP or HEC), the back-up generator, the uninterruptible power supply (UPS), the power distribution switchgears, the rack level power distribution unit (PDU) and the internal power supplies in each server equipment. When assessing redundancy capacity for data center/server room, the increase of the reliability of these critical power components will be included. Meanwhile, the backup power for data center/server room equipment can be reserved during a planned power shut down for maintenance. Power supply failures can be resulted from weather, equipment failures, power transmission/distribution line damages or a number of other causes. Any unplanned power failures can have detrimental effects on server equipment operation. At a minimum, a server system owner can incur significant financial losses and customer goodwill if it is unable to conduct business for any length of time after system server is out of order. Because of the substantial risks, reliable data centers/server rooms must provide sufficient redundancy to address any risk of unexpected electrical power failures. But redundancy comes in different forms. For example, data centers may offer power supply redundancy system. provide proper training and assistance to help customer during planning and implementing stages.

Solution

All data centers/server rooms will offer some form of redundancy. However, in most small or middle scale data centers/server rooms in commercial office building nowadays, alternative utility power supply (independent dual source) and back-up genset power supply is difficult to be provided by the landlord. Therefore, providing redundant design system in electricity supply main side is not feasible in most cases. In order to reinforce the electrical power system in sub-main side, 2N redundancy is proposed to maintain the same system reliability without alternative power and genset power supply scenario. 2N redundancy means the data center/server room has double the amount of equipment needed. The equipment runs separately with no single point of failure. This configuration represents a fully redundant system and a more reliable operation. Any power failure or planned power shut down will have no adverse effect on data center/server room availability. Applying a data center/server room with 2N redundancy is often required for mission critical environment. This type of capability is consistent with a Tier 4 data center – a facility which meets the strictest operational requirement. A Tier 4 facility installs two independent power paths. Each of these paths is capable of supporting the entire data center/server room with no single point of failure to disrupt system operation. If cost was not the main concern, all data center/server rooms should be considered to provide with 2N redundancy. Users can apply 2N under the most mission critical environment for those system operation not likely to be affected in a major way by an unplanned power failure. And the users will learn that the cost of downtime will far exceed the added cost of a 2N redundancy configuration. As from DCL’s data center operation & maintenance experience, zero downtime of IT system is recorded from 2N redundancy solution customers in Hong Kong since 2007.