Failures and recovery time

Change fail rate, failed deployment recovery time and deployment rework rate each rest on one kind of work item a Jira admin names in Settings: a failure. This page says what counts as one, and where Changeline looks. DORA used to call the first change failure rate and the second time to restore; incidents are no longer a measure of their own, an incident is a failure type.

A failure

A failure is a work item of a failure type that points at a deployment. The failure types are the work types named Bug, and every work type whose name contains Incident, which includes Jira Service Management’s incident type, by default. A Jira admin can name others.

• A failure points at a release when its Affects versions names the release.
• A failure also points at a release when it links to a work item of that release, with any link type.
• In Done mode, a failure is one that links to a deployed work item and was created after it.

A deployment with one failure or more failed. Change fail rate is the failed deployments out of all deployments, such as 3 of 25 releases. The Failures tab lists each failure, its work type, the release or work item it points at, the release that ships its fix, when it was created and its recovery time.

Where failures are read

Failures often live in a service space apart from the code’s, such as a Jira Service Management space. Under Failures in settings, Also read failures from these spaces adds them to the spaces of the scope. A failure there counts when it points at a release in the scope.

Failed deployment recovery time

For each failure recovered in the period, recovery time is its creation to the release that ships its fix, named in its Fix versions, else to its Done. The figure is the median, such as 6h 20m, with the count under it.

A failure not yet recovered is listed apart on the Failures tab, with Open in place of a time, and is named in the amber box above the tabs: 2 failures are not recovered yet: OPS-3, OPS-7. It joins the figure when its fix ships or it is Done.

Recovery time here is a work item’s creation to its fix, not an outage’s start to its end. It is as true as the failure work items and their Fix versions are kept.

Deployment rework rate

A deployment is rework when it ships a fix for a failure of an earlier deployment: it is named in the Fix versions of a failure that points at an earlier release. The rate is the rework deployments out of all deployments, such as 2 of 25 releases. The Releases tab has a Rework column.

DORA here is the research programme at Google behind these measures, not the EU’s Digital Operational Resilience Act. DORA is a trademark of Google LLC, and Changeline is not made or endorsed by Google.