What this check inspects
Shipwork builds the site map from the sitemap a bounded sample at a time, follows a sitemap index where there is one, and reads each listed URL. For every entry it records the status, whether the URL redirected, whether the page asks not to be indexed, and whether its canonical points somewhere else. It deduplicates the list, reports how many URLs were checked, and says when it stopped early to stay inside its time budget.
What a failure means
A listed URL that returns 404, returns a server error, or cannot be reached at all is an error, because the sitemap entry is not a working page. A listed URL that is set to noindex is also an error, since a page that asks not to be indexed must not be advertised in the sitemap. A listed URL that redirects, or that points its canonical at a different page, is a warning: the sitemap should list the final, canonical address rather than a detour. With no sitemap found at all, the check warns rather than errors, because a sitemap is optional.
How to fix it
- Remove entries that 404 or error, or correct the address if the page still exists.
- Replace a redirecting entry with the URL it lands on, so the sitemap lists final addresses only.
- Remove a noindexed URL from the sitemap, or remove the noindex if the page belongs in search.
- List the canonical address for any URL whose canonical currently points somewhere else.
A typical failure, worked through
The setupA store migrates its collection URLs and redirects the old addresses to the new ones, but the sitemap is regenerated from a cache that still holds the old handles.
What the check reportsThe check reports several sitemap URLs that redirect to the new addresses, and one that now 404s because the old collection was deleted. Every entry still loads for the team as a redirect.
The pointA redirect hides the stale entry from a browser and not from a crawler. The sitemap has to list the final address, which is what the check tests for.