Documentation
Overview
The dead link checker takes all links (a href) that are found within an HTML component and tries to open it. If the given resource is not reachable (HTTP status code >= 400) Leankoala will open an incident with all dead links.
Sometimes it is only important that all internal links work. External links can be neglected. That is why Leankoala provides an option to ignore broken links that point to URLs that are nor part of the same system. A link is internal if the domain equals the components domain. Subdomains are ignored.
The dead link checker does not only work with HTML pages, but can also check PDFs. This must be an unencrypted, text-based document. In this case Leankoala can extract all URLs and check their validity.
Sometimes links are allowed to be dead or they just act differently when retrieved by an automated tool like Leankoala. If those are single links they can easily be excluded.
The input field takes a string that must be part of the excluded URL. Those strings can look like:
example.com // exclude all links that point to example.com
example.com/help // exclude all links in help directory on example.com
robots.txt // exclude all robots.txt links no matter what domain
example.com/help // exclude all links in help directory on example.com
robots.txt // exclude all robots.txt links no matter what domain
Best Practice
Some patterns are excluded by default to prevent damage. Those patterns are among others: linkedin.com (always returns a 999), delete, create, remove.
The checker will also not react to SSH problems. You have to use the Https Certificate Checker for that.
Leankoala provides the possibility to log in into web applications. For those components, the broken link checker cannot be used. This is because the checker will "click" on every link that was found. So if there is a delete link, we would click it. That is why we do not provide this feature if the component is connected to a session.