A year ago, Microsoft promised to enable Webmasters a method of verifying MSNbot. Way too often, rogue spiders mask themselves as official spiders from Google, Yahoo, Live Search or Ask.com. The search engines have enabled methods to conduct reverse DNS lookups on the fly, so that you can allow those spiders that pass the reverse DNS test in and the others, deny at your gate.
I opened up my log files for rustybrick.com and found these records for MSNBot, Microsoft Live Search's spider.
65.54.165.35 www.rustybrick.com - [05/Dec/2007:01:44:46 -0500] "GET /robots.txt HTTP/1.0" 200 23 "-" "msnbot/1.0 (+http://search.msn.com/msnbot.htm)" "-"
65.54.165.35 www.rustybrick.com - [05/Dec/2007:01:44:47 -0500] "GET /seo_articles_8e.php HTTP/1.0" 200 11224 "-" "msnbot/1.0 (+http://search.msn.com/msnbot.htm)" "-"
Now if you do a reverse DNS lookup on this IP, 65.54.165.35, you will notice it returns the host name "by1sch4030208.phx.gbl" and not "livebot-[IP-Adress].search.live.com" as promised.
If you have implemented a reverse DNS lookup requirement for MSNBot, be warned that you may be blocking MSNBot from crawling your site, because Microsoft did not properly set up this IP block to reverse DNS to search.live.com.
The whois information for the IP confirms it is owned by Microsoft, but the reverse DNS does not.
jdMorgan, a WebmasterWorld moderator, said in the WebmasterWorld thread that there "are actually two PTR records for addresses in that range; If your server checks only the first (which appears to be a CNAME), then it may fail rDNS verification. The second record points to hotmail.com." But in any event, this is still a major issue for some webmasters.
Forum discussion at WebmasterWorld.
