Blogs

Proper Implementation of Redundant Power Supply at the Rack Level

Redundant Power Supply

When it comes to protecting data, there are many different important measures of precaution that need to be considered and implemented, and ultimately, ensuring your server’s safety and reliability means avoiding downtime at all costs. One of the most common provisions used to accomplish this is the redundant power supply, and it‘s one of the most commonly implemented technologies in data centers.

Redundant Power Supply

Redundant power supply contains multiple or “redundant” power modules within one power supply chassis so that in the event of some form of power failure, there is still a remaining power module(s) to keep the server from going down. Indeed, redundant power supply do a great deal of good for busy servers, but they can also be bad for the server if improperly installed.

Typically, a data center’s server administrator is responsible for managing the servers and power distribution units (PDUs). In the most common scenario, system administrators will install 2 rack-level PDUs and connect each of the redundant PSU outputs to a different PDU. This is okay if done properly, however it can lead to issues if not done right. For instance, if we are dealing with a dual or 1+1 redundant power supply that contains two power supply modules, when both modules are active, they share the load roughly 50/50. When one fails, the remaining module will begin to carry the full load. This is why it is very important to consider the max load capacity of your PDUs before implementation. If the PDUs attached are both at 60% of their maximum rated load and suddenly one of the power modules fail, the PDU is now 20% over capacity and you can end up crashing the whole servers. Unfortunately, this type of cascading failure is very common and while it may seem unlikely that a system administrator would invest in PDUs that cannot handle the full load of a redundant power supply, this incident is usually occurs as a result of the PDUs supporting other equipment on the rack.

The key with avoiding this problem is to never exceed 40% of your maximum load on all PDUs supporting dual or 1+1 redundant power supply. In fact there are UL and NEMA mandated codes that specify users can only draw up to 80% of a PDU’s load safely. So in summary, make sure that you thoroughly review your load structure and make sure that you are always monitoring your levels and analyzing their effectiveness, because you never know when a power issue may occur.

Leave a Reply

Your email address will not be published.