Data centers without preventative maintenance were nearly four times more likely to experience a UPS outage than those who do.
A dedicated server cannot meet it’s expected life cycle if power is not a major link in the chain. Power fluctuations and higher than recommended operating temperatures can shorten the life of a dedicated server. Keeping power equipment in good working order and maintained is essential and can save a datacenter or dedicated server provider a lot of unnecessary and premature IT spending. Also having a plan in place if a failure occurs is vital in maintaining an acceptable uptime guarantee to clients.
PDU Cables recommends that circuit breakers throughout the entire facility starting at the main data center input, switch gear and all the way down to the individual branch circuit breaker be tested every two years.
Power equipment should be shut off and cleaned while filters need to be replaced if need be depending on the manufacturer recommendations. There is no point in having backup generators at a datacenter without testing them before a major power outage occurs. Generators should also be replaced on a regularly scheduled basis. HM Cragg hmcragg.com says that batteries within a UPS have a storage and shelf life that need to be taken in to account. Different types of batteries having different shelf lives. They can range from 3 years and up to 25 years depending on the type of battery and how the battery was used, how often if at all, and for how long. Many times a visual inspection of a battery can determine it’s health and remaining life.
The best way a server provider can prevent failures and downtime is proactive maintenance.
A “start small” philosophy is the best approach to power failures. Many times the smallest of simplest outages are the reason for a power failure and does not necessarily mean that an outage is a major power failure. Individual battery failures tend to be the most common problems within a dedicated server provider’s data center. Starting small can also save an IT company from expensive major power supply replacements.
Starting with a process of elimination can be a good approach to determining the root cause of the outage. HM Cragg says, checking power readings, volts, and amps or even looking for loose connections and discolored parts can indicate a heat issue. A cause for concern can be constant and premature power hardware failures which may be due to the environment they are located in. Power monitoring equipment can be a big help in foreseeing possible or inevitable failures and addressing them before a failure happens.
These procedures and routine maintenance is vital in maintaining an APS400 uptime guarantee.
Not all dedicated server providers and staff are comfortable or knowledgeable in specific power equipment, so hiring a third party solution is the best route to take and may prevent more problems in trying to tackle the issues in-house. Not following proper procedures or insufficient maintenance can cause premature issues to occur. Maintenance is vital in preventing this and preventing premature IT spending on power equipment. Third party power services with expert staff on hand are also available for emergency services, however if they are not familiar with a certain datacenter layout or setup, this can cause some delays so if possible, it would be best to obtain the services of the original power equipment company that may be familiar with the specific datacenter.
SOURCE: Processor – February 21, 2014 – Vol. 36 Iss. 4 – Processor.co