A data centre is a network of computers, typically used by a large organisation or company. They can be responsible for storage, alongside processing or distributing the company’s data. It is essentially a physical location, where the hardware is kept — think of a room with corridors between a library of computer processes and network equipment. Or, these days, part (or all) of it could be on the cloud.
But occasionally, they fail and experience a data centre outage.
According to the Uptime Institute’s latest report, most organisations have experienced a data centre outage, which tends to happen for one of four reasons.
- Third-party services
- Network problems
- Power outages
- Hardware or software problems
Third-Party Services
This tends to be related to software or configuration problems. Companies cited particular concerns with cloud providers due to an over-reliance on one, or just a handful, of hyperscalers. Third-party issues may also be linked to network connectivity.
Network problems
Uptime Institute’s conceded that network issues are inevitable. Given network tools are so frequently updated, it’s bound to happen sooner or later. The issue lies particularly in the configuration management and malfunctions with third-party networks.
Power outages
Although among the most infrequent causes of a data centre outage, a power outage can be the most costly. In the report, less than a third of companies experienced a power outage-related data centre outage in the last three years. The recovery process was drawn out and expensive. The cause can be for many reasons, including as simple as a generator failure.
Hardware or software problems
The range of issues that fall under these banners is vast, especially if you have different data centres. But the most common cause of IT system problems was changes and updates. This is when you are most vulnerable to a data centre outage.
Cyber attacks were also a cause, though less frequent.
If the data is stored on the cloud, the outages are more frequent, though less severe. The ultimate conclusion from the report was that resilience to data centre outages relies on the competence of employees.