The data center reliability-sustainability paradox

Johanna Flood

Imagine never being able to rest. Ever. Not even a second. That is the reality of a data center.

Data centers need power all the time, 24/7. Only a few seconds of blackouts can lead to costly failures for both operators and users.

To ensure this, clients set strict uptime requirements in their contracts, often expressed as  99,9%. If these levels are not met, penalties follow.

To meet these perfectionistic demands, data centers have extra systems when one fails. Like two power lines, batteries, and backup power. Both hängslen och livrem as we say in Swedish (meaning both braces and belts, or to be on the safe side). The higher the up-time is for the data centers, the more extra equipment they have.

But this reliability comes at a cost. The more redundancy a data center has, the more equipment and materials it requires, potentially increasing its environmental footprint.

In the following articles, we will explore how this redundancy affects the environment and people, and if we can find the sweet spot between sustainability and reliability.

deep dive

The building blocks of data center and their sustainability risks

Best tips to select UPS for sustainability

UPS – the first aid for power failure

Servers are a bit picky. They need steady, clean electricity to work properly. All the time.

Best tips to select gensets for sustainability

Generators – intensive care during black out

If the power cut lasts longer, backup generators start up automatically shortly afterwards. Like intensive care for the data centers.

Author name

Cooling - keeping the servers from over-heating

Every watt that goes into a server eventually turns into heat. So a data center is basically a giant heat generator with some computing on the side. Cooling systems exist to remove that heat so everything keeps running instead of melting.