Skip to content
OPERATOR ACADEMY · GLOSSARY · DEFINED TERM

Concurrent maintainability

The ability to remove a planned component or path from service without interrupting the critical load.

Also called: Concurrently maintainable

In practice

Concurrent maintainability and fault tolerance are routinely treated as the same property, and they are not. Fault tolerance asks whether the load survives an unplanned failure. Concurrent maintainability asks whether planned work can be done — a filter change, a firmware update, a breaker inspection, a battery replacement — while the load keeps running. A design can have either without the other.

It matters more often than fault tolerance does, because planned work happens continuously and failures do not. A facility that cannot take equipment out of service without a service window will either defer maintenance until it causes a failure, or accumulate outages it scheduled itself. Both outcomes cost more than designing the maintenance path in at the start.

Achieving it means more than having a spare unit. The isolation has to exist — valves, breakers, bypasses — and the procedure to use it has to be written, approved and rehearsed. A maintenance bypass that nobody has operated in three years is a component with an unknown state, and the moment to discover that is not while the load is on it.

When someone describes a system as redundant, the question that separates the claim from the capability is simple: can you take that component out of service on purpose, on a Tuesday afternoon, without a service window? The answer is usually more informative than the redundancy label.

Scope of this definition

This is an educational summary of how the term is used in data center practice. It is not a standard, a specification or engineering advice, and where a real decision depends on it, the current adopted standards, verified site information and qualified professional review are the correct sources.

Back to the A–Z glossary · Operator Academy

Related material

Related terms

  • Single point of failure — Any component, path or shared dependency whose loss removes the service, regardless of how much redundant equipment surrounds it.
  • Fault domain — A set of resources expected to be affected by the same underlying failure.
  • MTTR — Mean time to repair or restore: the average time from a failure to the service being usable again.

Where this term is used

Reference