Concept · measure · registered

Failed deployment recovery time

How long it takes to restore service after a deployment, or a release to users, degrades the service and needs intervention. This is DORA's recovery metric, alongside deployment frequency, lead time for changes, and change failure rate. Until 2023 DORA asked for time to restore service after any user-impacting incident, which mixed failures caused by a change with failures from elsewhere, and that older question is not a synonym of this metric. A measured window is summarized by the median of the recovery durations. DORA's survey asks for the typical time rather than publishing a mean. The arithmetic mean of restore times for failures in general is mean time to restore. The clock ends when the service is restored, not when an alert is acknowledged and not merely when the incident record is closed.

coredraft@ 3b5f1abCC0-1.0

Definition

What Failed deployment recovery time means

Labels
en Failed deployment recovery time
Kindmeasure

Used in data

In OVDB databases

Databases, recordsets and fields whose meaning is this concept, a kind of it, or whose values are instances of it. Rows open the OVDB Directory.

No OVDB database carries this concept yet.

Measure

How it is computed

FormulaThe median, over failed deployments in the window, of the elapsed time from the deployment that caused the degradation until the service is restored. A failure that a deployment did not cause is excluded.
Aggregationnone

Start with meaning

Propose a definition

This local draft is for a semantic meaning only. Register logical data models or meaning files through the registry guide.

A local JSON draft downloads to your device. Nothing is published.