Connection health and Doctor
Read informational health factors and recorded failure guidance.
GET / reads health in the authenticated environment. The setup portal shows the same health to the organization’s IT admin. Health never changes authentication or policy.
Understand the observation
Section titled “Understand the observation”The response covers the preceding 24 hours and excludes test sign-ins. Below ten ordinary sign-ins, the overall score is omitted and the status is insufficient_data. Missing scores are not zero.
With sufficient evidence, status is healthy, degraded or unhealthy. The four factors describe:
signins: observed ordinary sign-in outcomes.certificate: the latest configured certificate expiry horizon.scim: an exact current active directory binding, without organization-based inference.standby: configured standby registration, not proof of successful failover.
Factor statuses distinguish observed, configured, unavailable, insufficient_data and not_applicable. Inspect the factors and observation time rather than treating a single score as a service guarantee.
Use Doctor for a failure
Section titled “Use Doctor for a failure”Doctor uses the recorded timeline and actual terminal validator evidence. Known error codes retain their catalogue guidance; historical records without newer evidence keep the supported earlier guidance. A timeline may include the cause, IT administrator message, steps and guide identifier, without assertion or attribute values.
See test sign-ins and Connection Doctor for testing and sending supported instructions, and sign-in history for ordinary attempts. Doctor explains a failure; it does not repair the IdP or bypass a check.
The health worker can notify eligible account owners/admins and subscribed organization IT contacts, with durable throttling and current recipient checks. Provider submission does not prove delivery. Customer IT sender configuration applies only to eligible IT mail; developer notifications remain platform-controlled. See customer IT email.