Quick troubleshooting view
The target is shown as DOWN and metrics are no longer updated.
- Unavailable /metrics endpoint
- DNS or network issue
- Authentication or TLS issue
- Open Status > Targets
- Read Last Error
- Test the URL from the Prometheus server
- Correct only the component confirmed by the checks
- Retest the original symptom after the change
- Escalate with collected evidence when the cause remains unclear
Contextual technician plan
Perform this check and preserve the observed result before changing configuration.
“Open Status > Targets” should produce an observation that clearly confirms or rules out “Unavailable /metrics endpoint”.
If the observation is normal, lower “Unavailable /metrics endpoint” in the ranking and continue with the next distinct check.
If the observation is abnormal, keep the evidence and investigate “Unavailable /metrics endpoint” first. Related action: Correct only the component confirmed by the checks.
Perform this check and preserve the observed result before changing configuration.
The exact failing name should resolve through the expected DNS server to the expected record without timeout.
If the exact name resolves correctly, compare application cache, suffix/search domain and the client or network where the failure remains.
If resolution fails or returns the wrong record, keep the queried server and answer and correct the resolver, zone/record or DNS path that is actually wrong.
Perform this check and preserve the observed result before changing configuration.
“Test the URL from the Prometheus server” should produce an observation that clearly confirms or rules out “Authentication or TLS issue”.
If the observation is normal, lower “Authentication or TLS issue” in the ranking and continue with the next distinct check.
If the observation is abnormal, keep the evidence and investigate “Authentication or TLS issue” first. Related action: Escalate with collected evidence when the cause remains unclear.
Repeat the same validation test after the correction and confirm the original symptom is gone. Validate stability before closing the incident.
Before changing configuration, record the current value and a way back.
Escalate with the exact symptom, scope, timestamp and completed checks when the issue remains unresolved.
＋Open the complete detailed guideDetailed explanations and original troubleshooting content.
The target is shown as DOWN and metrics are no longer updated.
Likely causes
- Unavailable /metrics endpoint
- DNS or network issue
- Authentication or TLS issue
- Timeout or oversized response
Checks in priority order
- Open Status > Targets
- Read Last Error
- Test the URL from the Prometheus server
- Check certificate, authentication, and timeout
When to escalate
Escalate when the failure affects multiple users, a production dependency is unavailable, or logs show a component outside your control. Include timestamps, scope, tests already performed, and the last known working state.