Skip to main content
Version: Next

Incident Age

What is this metric?

The amount of time it takes an incident to fix.

Why is it important?

  1. Help the team to establish an effective hierarchical response mechanism for incidents. Focus on the resolution of important problems in the backlog.
  2. Improve team's and individual's incident fixing efficiency. Identify good/to-be-improved practices that affect incident age

Which dashboard(s) does it exist in

How is it calculated?

Similar to requirement lead time, this metric equals resolution_date - created_date of issues in type "INCIDENT".

Data Sources Required

This metric relies on issues collected from Jira, GitHub, TAPD, or PagerDuty.

Transformation Rules Required

This metric relies on the 'type-incident' configuration in Jira, GitHub or TAPD's transformation rules while adding/editing a blueprint. This configuration tells DevLake what issues are incidents.

SQL Queries

The following SQL shows how to find the incident age of a specific incident.

-- lead_time_minutes is a pre-calculated field whose value equals 'resolution_date - created_date'
lead_time_minutes/1440 as incident_age_in_days
type = 'INCIDENT'

If you want to measure the mean incident age in the screenshot below, please run the following SQL in Grafana.

with _incidents as(
DATE_ADD(date(i.resolution_date), INTERVAL -DAY(date(i.resolution_date))+1 DAY) as time,
AVG(i.lead_time_minutes/1440) as issue_lead_time
FROM issues i
join board_issues bi on = bi.issue_id
join boards b on bi.board_id =
-- $board_id is a variable defined in Grafana's dashboard settings to filter out issues by boards in ($board_id)
and i.status = "DONE"
and i.type = 'INCIDENT'
and $__timeFilter(i.resolution_date)
-- the following condition will remove the month with incomplete data
and i.resolution_date >= DATE_ADD(DATE_ADD($__timeFrom(), INTERVAL -DAY($__timeFrom())+1 DAY), INTERVAL +1 MONTH)
group by 1

date_format(time,'%M %Y') as month,
issue_lead_time as "Mean Incident Age in Days"
FROM _incidents

How to improve?

  1. Observe the trend of incident age and locate the key reasons.
  2. Compare the age of incidents by severity levels, types (business, functional classification), affected components, etc.