The ReDock checks that visit your site
If you found ReDock-LinkCheck or ReDock-SafeUpdate in your access log, this page is what they are.
What it is
ReDock is a tool agencies use to look after the WordPress sites they build. One of the things it does is check each of those sites for broken links, so the agency finds them before a visitor does. The check is an ordinary visit to the public site: it reads pages and follows links, exactly as a person with a browser would, and it changes nothing and signs in to nothing.
What it fetches, and how often
- Roughly once a week per site, never more than that on a schedule.
- At most 20 pages and 150 link checks in one visit.
- One request at a time, with at least 300 milliseconds between requests to your server. Never two at once.
- It reads /robots.txt first and obeys it, including Crawl-delay.
- It requests HEAD where it can, and falls back to GET only where a server refuses HEAD.
The second check: around an update, not on a schedule
The other user agent, ReDock-SafeUpdate, is a safe update. When somebody who looks after this site asks ReDock to update its plugins, ReDock looks at the home page and up to 4 more pages the way a visitor would, runs the update, and then looks at the same pages again: if one of them comes back broken, ReDock puts the site back the way it was and tells the person who asked.
These requests arrive in a pair around a change a person asked for, never on a schedule, so you will see a short burst and then nothing until the next update. They read the pages and nothing else, and they sign in to nothing.
How to stop it
Add this to your robots.txt and the weekly check stops on the next run, with no argument from us:
User-agent: ReDock-LinkCheck Disallow: /
You can also allow it in part - any Disallow path is honoured, and so is a Crawl-delay if the default pace is still too fast for your server. If a site's robots.txt shuts us out, the agency who runs that site sees a line saying so rather than a clean report, so nothing goes unnoticed.
The safe update is not on a schedule and robots.txt does not govern it: it is part of a change the person who looks after this site asked for, and the way to stop it is to ask them not to update through ReDock.
If something is still wrong
Write to support@redock.xyz with the address of the site and roughly when, and we will look at it.