Geo-Redundancy

Last updated: 16.04.2026
Sie sind hier:
G
>
Geo-Redundancy

Geo-Redundancy

What is Geo-Redundancy?

Geo-Redundancy (geographic redundancy) describes a strategy in which IT infrastructure, data, and services are replicated across multiple geographically separate locations. If one location fails completely – due to a natural disaster, power outage, fire, or other event – the other locations continue operating without interruption. Geo-redundancy is the highest level of availability and disaster recovery planning.

Why geographic separation matters

Localised redundancy (e.g. two servers in the same server room) protects against hardware failures but not against location-wide events such as:

  • Natural disasters (floods, earthquakes, storms)
  • Power grid failures affecting an entire city or region
  • Fire in the building or data centre
  • Physical sabotage
  • Connectivity cuts affecting a data centre's internet connections

Only geographic distribution provides protection against these scenarios.

Minimum distances for geo-redundancy

There is no universally defined minimum distance, but industry best practices suggest:

  • Metropolitan redundancy: Two sites 10–50 km apart, connected with low-latency fibre links. Protects against most local events while allowing synchronous replication.
  • Regional redundancy: 100–300 km between sites. Protects against regional catastrophes; typically uses asynchronous replication due to latency.
  • Continental/global redundancy: Multiple regions across countries or continents. Highest protection level; suitable for global-critical applications.

Replication technologies

Synchronous replication

Every write operation is confirmed only after it has been written to all locations. This ensures zero data loss (RPO = 0) but introduces latency corresponding to the network round trip between sites. Practical with distances up to approximately 100 km.

Asynchronous replication

Write operations are confirmed immediately on the primary site and replicated to the secondary site with a short delay. Allows for greater distances but accepts a small data loss window (RPO > 0) in a worst-case scenario.

Geo-redundancy and cloud availability zones

Major cloud providers structure their infrastructure into Regions (geographic areas) and Availability Zones (independent data centres within a region). Using multiple Availability Zones provides metropolitan-level geo-redundancy; using multiple Regions provides regional or global redundancy.

Geo-redundancy requirements in regulated industries

The German Federal Office for Information Security (BSI) and the European Banking Authority (EBA) publish guidelines that specify geo-redundancy requirements for critical infrastructure and financial institutions. ISO 27001 and BSI-Grundschutz also address geographic distribution as part of business continuity planning.

Costs and trade-offs

Geo-redundancy involves significant costs: duplicate infrastructure, replication bandwidth, and potentially more complex application architectures. However, for organisations where downtime causes regulatory penalties, reputational damage, or direct financial loss, the investment is justified by the risk reduction.