BlackboxExporterProbeScrapeFailed

Prometheus is not getting a response from blackbox_exporter for a probe job, so there is no probe result to evaluate.


Severity	warning
Source	Prometheus scrape of blackbox_exporter `/probe`
Key metric	`up` for the blackbox job

What it means

For probe jobs, each “target” in Prometheus is really a request to the exporter’s /probe endpoint. up reflects whether that HTTP request to the exporter worked, not whether the probed endpoint is healthy (that is probe_success). The alert fires when those scrapes keep failing for several minutes.

While it fires, probe_success for the affected targets goes stale, so probe alerts cannot fire either. You have a monitoring blind spot rather than a confirmed outage.

Common causes

blackbox_exporter pod down, crashlooping or OOM-killed.
Scrape timeout: the probe takes longer than scrape_timeout, so Prometheus gives up before the exporter answers.
Relabeling mistakes, so __address__ no longer points at the exporter.
Unknown module name in the module parameter, which makes the exporter return an error.
Network policy or Service change blocking Prometheus from port 9115.

First checks

Find the failing scrapes and the error text in Status > Targets, or query:
```
up{job=~".*blackbox.*"} == 0
```

Check the exporter is running and its config loaded:

kubectl -n monitoring get pods -l app.kubernetes.io/name=prometheus-blackbox-exporter
curl -s http://<blackbox-exporter>:9115/metrics | grep blackbox_exporter_config_last_reload_successful

Call the probe exactly as Prometheus does and time it:
```
time curl -s "http://<blackbox-exporter>:9115/probe?target=<target>&module=<module>&debug=true" | tail -n 20
```
An error about the module means it is missing from the exporter config.
Compare the probe duration with the job’s scrape_timeout:
```
scrape_duration_seconds{job=~".*blackbox.*"}
```

Fixing it

Restore the exporter (resources, config, image). Fix relabeling so __address__ is the exporter and __param_target the endpoint. Keep the module timeout below the scrape timeout; the exporter also honours the timeout header Prometheus sends, reduced by --timeout-offset.

BlackboxProbeFailed: the probe runs but the endpoint fails.
BlackboxSlowProbe: slow probes that approach the scrape timeout.
BlackboxSslCertificateWillExpireSoon: expiry data stops updating while scrapes fail.