How to determine when a standby generator is needed in small data centre environments
At on365 we know that small data centres deploy Universal Power Supplies (UPS) in a wide variety of circumstances. We are able to provide a rational framework for establishing backup time requirements and how to determine the tradeoffs between supplemental batteries and a replacement standby generator. Our framework is not based simply on the latest and greatest technology but more on a detailed TCO analysis to establish which solution makes the most financial sense.
Virtually all small data centres and network rooms in the UK have existing UPS protection however the runtime of these devices varies dramatically across an infinite number of different installation and architecures. For example, a typical UPS installed in a network room may have a fullload runtime of only twenty minutes, whereas the UPS supporting your VOIP LanPBX may have a runtime of over 4 hours. It is quite routine to deploy a variety of differing strategies to provide additional runtime. These strategies include, but are not limited to:
on365 can help understand exactly how much runtime is appropriate or necessary for your specific application, and how much additional availability you can expect when runtime is extended. We bring a rational approach to assessing downtime risks and ensuring that required runtimes are met. This approach is comprised of four key steps:
Step 1: Establishing and avoiding the consequences of data centre downtime
Knowing the effects of backup time exhaustion is an important part of calculating the value of additional runtime. These consequences may include:
Step 2: Identifying specific runtime requirements
When considering back-up time requirements it is imperative for the system specifier to consult with key end-user personnel so that there is a clear understanding of the expectations for system performance. The two factors to consider in this discussion are the business costs of unplanned downtime and the risk profile for the site.
Step 3: Maintaining Cooling
In a powercut situation it is essential to continue providing cooling to the data centre as well as power. Any device running on substitute power but not being cooled is prone to overheating and eventual catastrophic failure. In data centres heavily populated with blade servers and other high demand devices, thermal overload can occur within 5 minutes, meaning it’s extremely important to ensure that all devices run within the manufacturers operating temperature guidelines. For small data centres or even well populated wiring closets, battery-backed fan ventilation may be sufficient to cope with these extremes during extended outages. The basic concept is to conduct the hot air out of the data environment, mixing with the ambient temperature in the building to ensure that the wiring infrastructure never exceeds this level.
Step 4: Choosing a Generator or an enhancement to UPS facilities
Enhancing the existing UPS infrastructure is reasonably simple and indeed can sometimes be the only option where there’s insufficient space for a generator. However if extended runtime is necessary and high density loads are normal then the only logical option to keep the devices running effectively is a generator. A UPS enhancement is likely to be more cost effective that a generator in many smaller data centre installs however there is often a “tipping point” where the added performance of a generator overcomes the economic advantag
Extended backup time on existing UPS devices can be a cost effective way to increase uptime in network rooms however these same backup runtime requirements can be a primary driver of UPS system cost. on365 can help your business estimate the cost of downtime, understand the downtime probability of your infrastructure and ultimately determine the exact runtime requirements to protect your investment. Once these basic parameters are understood it becomes clear at what point you need to upgrade to a standby generator.