Server response check

The server response check is the most-used diagnostic in the console, because it shows what the robot receives rather than what a browser renders. Yandex opens with the rule that makes it decisive: if the response is not 200 OK, or carries one of the listed messages, the page is excluded from search results.

What it reports

• The HTTP status.
• The SSL certificate status for HTTPS sites — with a long list of specific faults: issuer not found, signature invalid or undecryptable, certificate not yet active, expired, self-signed and untrusted, chain containing only one non-self-signed certificate, revoked, chain too long, wrong certificate type, root untrusted or blacklisted, or a certificate already issued to another site. These are resolved with the certificate authority or hosting provider, not on the site.
• Diagnostic messages covering the common failures.

The messages worth knowing

• Document failed to load / failed to connect. Three causes: the robot blocked at the server (check IP blocking), a DNS error, or a timeout — the response must arrive within 3 seconds.
• noindex tag disables indexing, or indexing disallowed in robots.txt.
• Document too large — the 10 MB limit.
• Unsupported format — HTML plus PDF, DOC/DOCX, XLS/XLSX, PPT/PPTX, ODS/ODP/ODT/ODG, RTF, TXT and SWF.
• Language not recognised, or document contains no text.
• Page content is missing — check the HTTP header, because Content-length: 0 makes the page unindexable, and the declared size must match the actual content.

How to use it properly

Select the specific robot — the dropdown carries the main agents, and testing only the default hides mobile-specific blocking. Use the If-Modified-Since field to confirm conditional responses: the page’s real modification date should return 200, any other date 304.

The caveat Yandex states itself

The tool requests from a different IP than the robot, so its result may differ from what the robot actually receives. When the two disagree, the server log is the authority.