The sitemap error reference splits into two groups: failures to fetch the file, and failures to parse it. The distinction matters, because the first group is a server problem and the second is a generator problem, and they are handled by different people.
Download failures
Not indexed (download failed, URL possibly wrong), HTTP error (a status other than 200), invalid URL, connection lost, connection failed, DNS error, incorrect HTTP response, an empty response, a gzip extraction error, and URL forbidden in robots.txt — the last being an entirely self-inflicted case worth checking first.
The hard limits
- 50 MB maximum file size.
- 50,000 URLs per sitemap file — beyond that, split the file and use an index file.
- 50,000 sitemap files per index file.
- 1,024 characters maximum URL length.
- 100 errors and processing stops entirely.
The rule most generators get wrong
URL does not match the sitemap file location. A sitemap in a directory may only contain URLs from that directory or its subdirectories. A file at /sitemaps/products.xml listing URLs at the site root is invalid, and this is exactly what many plugins produce when a sitemap is moved out of the root.
Encoding and structure
The file must begin with the UTF-8 prefix and the standard XML declaration; missing it is a warning (“encoding not UTF-8”) rather than an error. Structural errors cover tags appearing more than once, unexpected tags, missing required tags, empty tags, nested tags where none are allowed, and tags with too much data. Invalid XML — a file that is not well-formed — stops everything.
Two early-termination conditions
Yandex stops processing when it sees more than 100 errors, and also when a file begins with consecutive invalid URLs, on the assumption that the format is wrong. Both mean the reported error list is incomplete, so fixing what is shown and resubmitting can reveal an entirely new set.
Nesting
A sitemap index file may reference sitemap files only — not other index files. Deeply nested structures on large sites fail here.
The warning to ignore
“Unknown tag (ignored)” is often a legitimate sitemap protocol extension for another engine. Yandex ignores the tag and processes the rest; the file does not need changing.
Comment nous l'appliquons
The error we see most often on client sites is the location rule — a plugin writing a sitemap into a subdirectory while listing root URLs — and the second is a 50,000-URL overflow on a catalogue that grew past the threshold without anyone noticing. Both produce a file that looks perfectly reasonable when opened in a browser. We check the sitemap after every catalogue expansion for exactly that reason.
Services associés