I’ve got a few Facebook Applications I’ve played around with developing that are not actually for use (read: they do nothing). I’ve noticed over the past few days their canvas URL’s are seeing traffic in the form of 1 hit approximately every 24 hours. Previously they saw no traffic at all. At first I thought this was just Facebook with some new process to check for malicious apps, which sounds like a good idea. Then I did some digging and found something surprising:
The first thing I found was the hostname where the request originated was
out-sw251.tfbnw.net which is obviously owned by Facebook. That’s not terribly interesting and supports my theory up above.
Then I found these two curious bits in the request:
X-FB-USER-REMOTE-ADDR: 126.96.36.199 USER-AGENT: Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)
That IP address is
crawl-66-249-67-211.googlebot.com. That UserAgent is very telling and needs no introduction.
The request is otherwise pretty unremarkable other than no query string which a normal person would generate when hitting that canvas URL. However
fb_sig_request_method is set to
GET which suggests to me it’s actually using
POST despite that what it claims. There’s no
fb_sig_user or anything else that would suggest an actual user, which makes sense because
fb_sig_logged_out_facebook is set to
It appears as of March 20, 2011 Google has started crawling Facebook Apps. I’ve got no idea what it’s intent, abilities or relationship is. I can tell you that I’ve monitored since at least April 2010 and this only started a few days ago.