How does googlebot know URLs I'm visiting in my browser??
Today I've been really struck at what I discovered and I can't find any explanation. I am working on an online shop and there is a mechanism which enables clients to confirm their email address by clicking a personalized link which they receive in an automatically generated message. Here is an example:
Of course, the link above is invalid but it's not the point. When a client clicks the link then the PHP script on the server marks this order as confirmed in the database. While I was testing this I was startled by what I saw in the server logs. So this is what happens when I go to the confirmation URL in my browser:
1. First page reguest is http://www.gsm-support.net/en/confir...806e33e2193232 - my IP has been logged, all is as expected
2. A second later comes another request for the same page and it was not made by my browser but by googlebot! The IP is 184.108.40.206, host name crawl-66-249-71-59.googlebot.com, user agent: Mediapartners-Google.
3. In the same second comes another same request from googlebot.
The script is therefore invoked 3 times. How can this be explained? I tried copying and pasting the link to Seamonkey, Firefox, Opera and IE and the result was the same every time. The only difference was IE in that there was only 1 request from googlebot. It's not possible that googlebot could obtain the link from other sources because it is generated at the time of invoking the script. And of course, I want to make sure that the link is known only to the person receiving it whereas it turns out googlebot knows it immediately! Could it be that my computer has some malware and sends to google URLs of the pages I visit? Or could the server where the shop is hosted be hacked and it is sending the information?
Also I observed that this happens only when I first visit the confirmation page. On subsequent visits only the request from my browser is logged. But when I make new order and new link is sent to me it happens again. Googlebot doesn't go to this URL first, it always goes there immediately after me.
I don't know if it's important but the confirmation email is sent via PHP using SwiftMailer without using my own smtp - in other words SwiftMailer in effect uses the php mail() function.