Not for an ordinary site. A scan asks for each page in your sitemap once, a handful at a time, and names itself as LinkSignalsBot so the visits are easy to recognise in your logs. For a single site that means roughly one request per page listed, spread over the length of the scan.
If even that is too much, you can block the crawler or ask for your domain to be excluded.
What a scan asks of your server
| Request | How often |
|---|---|
| The sitemap files | Once per scan. |
| Each page the sitemap lists | Once per page, up to the number of pages your plan reads. Up to six requests run at once during a scan, spread across every site involved, so your server is not the only one being asked. |
| The same page as a search crawler sees it | On the homepage and about 10% of pages: two extra requests each, to compare what a crawler is shown. This is how cloaked links are found. |
| A retry | Only if a request returns an error, and then once, with a browser-style user agent. |
Each page is given ten seconds to answer and about 1.5 MB is read, so a slow page does not hold anything up for long. The links on your pages lead to other sites, and those sites receive the requests for them, not yours. Links within your own domain are not followed.
What monitoring asks of it
The first monitoring check reads the whole site, in stages if the site is large. After that, each check reads new pages, pages whose sitemap date changed, and a rotating slice of about 5% of the rest (never more than 2,000 pages). A full pass happens about every 30 days. Monitoring is therefore lighter than repeating a scan, and the load drops further when your sitemap dates are accurate. Checking your sitemap dates shows how to tell.
How often it visits
A one-off scan visits only when someone runs one for your site, and there is no schedule. Monitoring visits on the schedule chosen for the site, and a manual check can be queued once an hour per site. There is more in how monitoring works.
If you want it to stay away
- Ask for your domain to be excluded. Nothing is then requested from it, whether through a scan of your sitemap or the check of a link that points at your site.
- Block the
LinkSignalsBotuser agent on your server. This stops scans of your own pages. A link to your site in someone else’s scan is retried once with a browser-style request, so the block alone does not stop that. Disallowrules inrobots.txtdo not stop it. That file is read only to find sitemap addresses.
Blocking Link Signals has no effect on how search engines see your site, because it is not a search engine. The rules for nginx, Apache and Cloudflare are in blocking the Link Signals crawler.
Where to read more
- Sitemap discovery: the timeouts, size limits and redirect handling.
- Does Link Signals need access to my site?: why nothing is installed or changed.
- What if my site is very large?: how big sitemaps are read.
- Google on robots.txt: what robots.txt rules do for search crawlers.
Questions
Can I see the visits in my server logs?
Yes. Requests carry the user agent LinkSignalsBot with a link to the Link Signals site, so they can be filtered or counted separately from other traffic.
Why does a scan take longer than my server needs?
Reading your pages is only part of it. Every link to another site is also checked, and slow or blocking destination sites stretch a scan out. That is why larger scans are given a longer time estimate.
Can I schedule scans for a quiet hour?
A one-off scan runs when it is started. If the timing matters, start it at a quiet time. Monitoring runs on its own schedule and its first full pass can take several checks on a big site.
Run a scan on your own sitemap and watch the page count against your server’s logs.
