IncidentSeoUrlSweepFailing

Severity: warning · Fires after: 13h · Rule: central/prometheus/rules/visibility.yml

What it means

For three polls running, a site's URL-Inspection batch either could not start (its sitemap could not be read, or was empty) or recorded nothing (every inspection call was rejected). Google index coverage for that site is not being measured.

Expression

webmaster_url_sweep_up == 0

webmaster_sitemap_urls tells the two causes apart: 0 = sitemap problem, > 0 = inspections failing.

First checks

ssh jakefear@docker2 'cd /opt/jakemon && docker compose logs --since 13h visibility-exporter | grep <site>'
curl -sv -o /dev/null https://<site>/sitemap.xml 2>&1 | grep -iE "HTTP/|SSL|certificate|subject"

How to clear

• Sitemap unreadable — fix the site. A TLS failure here (found on thelocalwiki.com, 2026-10-01: certificate hostname mismatch / expired as seen from docker2) is a site problem crawlers hit too, not a monitoring one.
• Inspections rejected (HTTP 403) — the Google account behind GOOGLE_REFRESH_TOKEN is not a verified owner of that sc-domain: property. Verify it in Search Console, or drop the site's google entry from visibility/sites.json.

Notes

Before this metric the only trace was a log line per poll. The exporter deliberately abandons a batch after 5 consecutive failures so a dead property does not burn quota.