IncidentSeoGoogleServerErrors

Severity: warning · Fires after: 6h · Rule: central/prometheus/rules/visibility.yml

What it means

URL Inspection reports one or more of a site's sitemap URLs in coverage state "Server error (5xx)": the last time Google crawled them, it got a 5xx. Google drops URLs that keep failing.

Expression

webmaster_url_coverage{state="Server error (5xx)"} > 0

Read this before acting

This is Google's record of its last crawl, not the site's current state. It only changes when Google recrawls, so it stays non-zero for days or weeks after the origin is fixed. First seen 2026-09-12 (20 URLs, then 68) following the 20-day transit outage.

First checks

# Is the origin healthy NOW, as a crawler sees it?
curl -s -o /dev/null -w '%{http_code}\n' -A 'Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)' https://<site>/
# What is the edge actually serving crawlers?
curl -sG 'http://192.168.0.5:9090/api/v1/query' --data-urlencode 'query=sum by (bot,status) (cloudflare_bot_requests_status{bot="googlebot"})'

How to clear

If the origin is healthy and Googlebot is being served 200s, the count is residue: use Search Console's "Validate fix" on the Server error report, then silence this alert with an expiry a few weeks out rather than letting it re-notify. If the count is growing, the errors are current — check SearchCrawlerErrors, the cloudflared journal, and up{job="wikantik"}.