What is a server farm?

(Almost) every human on the planet creates data. Including us (yes, we’re human). Including you. How? By using social media (and how!), taking photos and videos, using workplace software, etc. How does all this data get from your phone out in the world? Through servers. A server is a computer designed to process requests and deliver data to another computer over the internet or a local network. A lot of data means a lot of servers, a lot of servers mean a lot of server farms, and a lot of server farms mean a lot of maintenance. It’s very important for servers to be up and running at all times. The cost of server downtime is huge, somewhere between 301,000 and 400,000 U.S. dollars. We told you it’s huge!

The core needs of server farms

Oh, and a lot of computers mean a lot of energy consumed. And server farms have a lot of computers. The electricity used by them is ultimately converted into heat, which must be removed by cooling equipment that also runs on electricity. So the power usage is very important for server farms. In order to learn how effective one is, the total power requirements of computing equipment should be compared against its total power consumption (this is PUE - Power Usage Effectiveness). High PUE means inefficient farms which usually pass their high energy and cooling costs on to their customers, while low PUE means a more efficient farm that can deploy more powerful computing and network solutions at a far more competitive price point. That is why it’s crucial to prioritize optimization. Any optimization strategy must begin with data. Collecting information about server power consumption, and cooling performance establishes important baselines for measuring the efficiency of an infrastructure. No need to mention that Sensix does that, right? OK, let’s move on.

OK, track the data, but what data?

The first and most important data that should be monitored is energy consumption. Usually, the energy consumption is measured as an overall value. Our solution helps managers to track energy consumption down to the rack cluster, to have a granular overview of the energy usage.

Energy consumption is a very important metric when discussing optimization. But it is important to also measure the power quality. Power quality is used to describe electric power that drives an electrical device and the device’s ability to function properly with that electric power. Without the proper power, an electrical device may malfunction, fail prematurely, or not operate at all. For data centers and their clients, this means huge losses.

Another vital factor to monitor is temperature. Server rooms contain a mix of both hot and cool air – server fans push out hot air while running, while air conditioning and other cooling systems bring in cool air to counteract all the hot exhaust air. Maintaining the right balance between hot and cool air has always been foremost in maintaining data center uptime. If it gets too hot, equipment runs a higher risk of failure. That failure often leads to downtime, data loss, and profit loss.

Humidity is also essential. Ambient cooling always creates humidity in the air; it’s very important to make sure the humidity stays in the recommended range. If humidity is too low, the dry air will lead to electrostatic discharge which can damage critical server components. Too much humidity will cause condensation, leading to hardware corrosion and equipment failure.

A solution? Ours!

Having proactive monitoring equipment that can quickly and in real-time notify you when an environmental issue occurs is critical. Sensix offers an analytics platform, to compare historical trends, understand patterns, get specific recommendations so that you can better manage your data center, with a clear understanding of the ambient and energy. Learn more and get in touch.