How to protect your e-business from costly outages?

What does one hour of downtime mean for your organisation? How about one day? What would happen to deadlines and what would be the economic damage? In the modern world, these traumas are caused by failures in IT devices or data movement which have evolved into the lifeblood of organisations. Everything starts and ends with information and its movement. However, there is a cure for such interruptions if you keep your IT devices in a place where their reliability is guaranteed in more than one way.

Cooling systems of the data center. Photo: Tõnu Tunnel

How to survive?

These days, many organisations continue to keep their valuable data and IT systems in an office, basement, or some other room that has been converted into a server room. Understandably, these places are often as exposed to interruptions as any other ordinary building. However, the situation is different in data centers designed for the impeccable operation of IT devices.

Last month, we wrote about the high-tech security systems of our data center. We prioritise reliability as well as safety – what is the purpose of state-of-the-art security systems if the infrastructure fails?

Greenergy Data Centers is the first colocation data center in Estonia to have a detailed plan and building permit tailored to the needs of the data center.

Where is reliability born?

The keyword here is the redundancy of support systems. This suggests that the system includes extra components, more than the minimum requires. Although, oftentimes, redundant things can be a nuisance, it is definitely not the case with business-critical infrastructure. There simply must be failsafes.

Take an average office building as an example, where many keep their devices – everything there is non-redundant. To put it simply: if somebody digs through the internet cable during roadworks, you have no internet connection. Or if a cat jumps on the transformer at the nearest substation, you have no electricity. In other words, it does not take much to interrupt work.

It is these moments that redundancy becomes a lifeline for the data and activities that rely on it. If the data center is connected to the network by communication cables from eight different directions and to one of the centremost electricity nodes by two independent connections, which are in turn supported by UPSes and electricity generators, then there is something to count on. This will rule out a situation where the careless work of a single excavator driver could jeopardise the whole operation and other such incidents.

So, in this case, redundancy does not mean excessiveness, but being better prepared, which ensures the smooth operation of IT devices even in worst-case scenarios. This is the most straight-forward way of preventing problems.

How does it work exactly?

The reliability of Greenergy Data Centers is ensured by a redundant electricity supply, cooling, and infrastructure, state-of-the-art fire prevention systems, and the continuous monitoring of the center’s infrastructure with more than 3000 sensors.

The electricity supplied to the center’s server rooms is circulated through uninterruptible power supply or UPSes that ensure power, if, for some reason, normal power supply fails. UPSes are meant to last until the standby generators placed in marine containers, which are able to generate electricity for as long as needed, are operational. Large fuel tanks have been built underground in the premises of the center, meaning that the data center can run freely on fuel only. This ensures power even when the lights are out in homes and offices.

The same is true of the cooling systems of IT devices, which are built on the logic that both the shared data hall and the private rooms all have back-up equipment that become operational if any of the cooling units which keep the temperature within the appropriate range stop working. The rated power of the largest cooling units at the Greenergy Data Centers is 1.5 megawatts. When seeking comparisons from everyday life, there is a 100,000-fold power difference if we take a regular mobile charger (15 watts) as an example. If the cooling modules were to stop working for some reason, there are large tanks with cold water on hold, the contents of which can be used in the cooling pipes until the temperature control devices restart.

As a fire safety measure, the oxygen level in the data rooms has been reduced to prevent fire. Figuratively, this means that if you walk into a data room with a burning torch, the torch will go out on its own. For a person in good health, this environment is suitable for a couple of hours of work. This can be compared to being in the mountains at a height of a few kilometres.

Cleverly selected location

According to international standards, there are more than 100 different criteria for selecting the location of a data center. Data should be kept away from dangerous traffic and manufacturing. It should still be close enough to main transport hubs. It is wise to build the center higher, since this reduces the risk of flooding and increase the number of days with good winds. This way, the colocation data center can be cooled for most of the year with outdoor air, which saves energy. The Estonian climate is perfect for such a facility and there are only a few hundred hours a year where free air cooling is not enough. By the way, Estonia is also among the ten most suitable countries for building a data center.

How much does downtime cost worldwide?

According to an annual survey by Information Technology Industry Council (ITIC), 98% of companies claim that an hour of downtime will cost them over 100,000 dollars (about 94,000 euros). Downtime may also affect employee productivity, as they must deal with systems and data recovery and cannot focus on their main work that would move the business forward.

According to a study by Vertiv – an American company providing data center devices and services – unplanned downtime might end up costing hundreds of thousands of euros to a company: in case of partial downtime, about 243,000 euros on average, and in case of a full blackout, about 642,000 euros. The hefty price tag is related to the duration of the outage and the size of the company. The main (and most costly) reasons for outages were directly related to the vulnerabilities of power supply and cooling infrastructures of IT devices.

All in all, it is worth asking your data center, cloud services provider, or office equipment specialist – what will happen in the event of an unexpected incident or an accident? Is there a plan B or even a plan C? At the end of the day, instead of summing up the damage and putting out fires, we all wish to look to the future with peace of mind.