Analyze service health in Service Observability
Use the Analyze Service Health skill to view an analysis of all Service Observability dashboards for the selected service.
About this task
The analyze service health skill uses generative AI to analyze all available Service Observability dashboards for the selected service.
When you ask the skill to analyze a service's health, it generates an optimized version of the Service Observability charts for a service, whether or not they are currently displayed. It then analyzes the data, and taking the time period shown into account, detects reports any insights it finds, and analyzes general service health. With these insights on service health, operators can assess service impact, identify root causes, and streamline incident response.
The skill's results remain available for an hour and don't automatically refresh. You need to rerun the skill to view a new analysis.
Once the skill is activated, it automatically runs for you in incident investigation from the Investigate tab in Incident Management. See Digital End-User Experience and Service Observability UI experience on investigate tab for more information.
- The analysis captures the time range as reflected by the user's set time zone. When sharing analysis data across time zones, be aware that the analysis time period always reflects the time zone originally captured and may not be in sync with your time zone settings.
- The analysis doesn't weigh insights with regards to possible impact to an alert, incident, or overall dashboard health.
- The skill has a maximum number of 50 images that can be analyzed.
Before you begin
You need to manually activate this skill before it can be used. See Activate the analyze service health skill for more information.
Role required: admin, operator