What is a Tier III datacenter, in practical terms?

At Hodi, we only host your data in Africa in Tier III datacenters, never below. But what does this level actually cover, and what does it change for you? Here's the explanation.

Tier III is a datacenter classification level defined by the Uptime Institute, the reference body on the subject. Its main characteristic: concurrent maintainability. In other words, any component (power, cooling, network) can be worked on without ever cutting the service you host. It's this property, not an uptime figure, that defines Tier III.

What "Tier III" means in practice

The specificity of Tier III comes down to one word: N+1 redundancy. For every critical function (power, cooling, connectivity), there is at least one backup element ready to take over.

In practice, when our teams need to replace a UPS unit, switch a power source, or carry out planned maintenance, they do it on the backup path while the primary path keeps serving your applications. You don't have a downtime window to plan on your end, and your users don't notice anything. This is the practical difference from Tier II, where maintenance on the distribution path does require a cutover.

What keeps uptime holding

These principles translate into specific equipment. Here's what keeps the service online day to day:

Dual network and dual power supply: two independent paths, one takes over if the other fails.
Duplicated backup generators and UPS units: autonomy if the power grid goes down, and no interruption when switching to backup.
Redundant cooling: air conditioning doesn't stop because one unit fails.
Multi-carrier connectivity: losing one carrier doesn't cut access to your services.
Maintenance without service interruption: the direct consequence of everything above.

It's this stack of redundancies that makes our 99.9% uptime commitment realistic over the long term.

Physical and environmental security

The Tier classification doesn't cover this layer as such. It rates power and cooling, not access security or fire safety. Our datacenters add it because a secure facility isn't just about uptime:

Video surveillance and 24/7 human presence
Secure access via badges and biometrics
6 physical doors to pass through before reaching the servers
Fire detection and waterless suppression, to put out fires without flooding the machines
Continuous environmental monitoring: temperature, humidity, air quality

Why Tier III is our foundation

It's the standard we can maintain identically in every African country we cover. Aiming higher where it isn't available everywhere would fragment the offering: one quality level in Senegal, another in Cameroon, another in Madagascar. By building on Tier III, we ensure a consistent, secure network, with the same quality of service everywhere. It's this common foundation that makes our commitment sustainable from one country to another.

Our environments are built on Tier III datacenters in every country we cover, with the same quality of service everywhere.

One point of honesty to close: our uptime commitment is contractual, it's ours. It shouldn't be confused with the Tier III classification, which describes an architecture, not a percentage. In fact, the Uptime Institute removed any reference to an annual downtime figure from its standard back in 2009 (official explanation). The two go together: it's because the infrastructure is designed for maintenance without downtime that this commitment holds up over time.

To learn more

Updated on: 17/07/2026

Was this article helpful?

Thank you!