DNS poisoning at Google?

David Miller dmiller at tiggee.com
Wed Jun 27 05:18:38 UTC 2012


On 6/27/2012 1:13 AM, Matthew Black wrote:
> I'm not familiar with curl and don't understand what I type and what are results. Are you suggesting that when google refers to our website, we pick that up and redirect to couchtarts?
>
> matthew black
> information technology services
> california state university, long beach

Referer is an HTTP header that can be included in requests to your web
server
  - http://en.wikipedia.org/wiki/HTTP_referer

"man curl"

       -e, --referer <URL>
              (HTTP)  Sends the "Referer Page" information to the HTTP
server. This can also be set with the -H, --header flag of course.  When
used
              with -L, --location you can append ";auto" to the
--referer URL to make curl automatically set the previous  URL  when 
it  follows  a
              Location: header. The ";auto" string can be used alone,
even if you don't set an initial --referer.


$ curl -v -e 'http://google.com' csulb.edu
* About to connect() to csulb.edu port 80 (#0)
*   Trying 134.139.1.60...
* connected
* Connected to csulb.edu (134.139.1.60) port 80 (#0)
> GET / HTTP/1.1
> User-Agent: curl/7.24.0 (x86_64-pc-linux-gnu) libcurl/7.24.0
OpenSSL/1.0.0g zlib/1.2.5
> Host: csulb.edu
> Accept: */*
> Referer: http://google.com
>
< HTTP/1.1 301 Moved Permanently
< Date: Wed, 27 Jun 2012 05:11:39 GMT
< Server: Apache/2.0.63
< Location: http://www.couchtarts.com/media.php
< Content-Length: 243
< Connection: close
< Content-Type: text/html; charset=iso-8859-1
<
<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>301 Moved Permanently</title>
</head><body>
<h1>Moved Permanently</h1>
<p>The document has moved <a
href="http://www.couchtarts.com/media.php">here</a>.</p>
</body></html>
* Closing connection #0


-DMM

>
>
>
>
> -----Original Message-----
> From: Jeremy Hanmer [mailto:jeremy at hq.newdream.net] 
> Sent: Tuesday, June 26, 2012 9:59 PM
> To: Matthew Black
> Cc: nanog at nanog.org
> Subject: Re: DNS poisoning at Google?
>
> It's not DNS.  If you're sure there's no htaccess files in place, check your content (even that stored in a database) for anything that might be altering data based on referrer.  This simple test shows what I mean:
>
> Airy:~ user$ curl -e 'http://google.com' csulb.edu <!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN"> <html><head>
> <title>301 Moved Permanently</title>
> </head><body>
> <h1>Moved Permanently</h1>
> <p>The document has moved <a href="http://www.couchtarts.com/media.php">here</a>.</p>
> </body></html>
>
> Running curl without the -e argument gives the proper site contents.  
>
> On Jun 26, 2012, at 9:35 PM, Matthew Black <Matthew.Black at csulb.edu> wrote:
>
>> Yes, we've used the Google Webmaster Tools a lot today. Submitted multiple requests and they keep insisting that our site issues a redirect. Unable to duplicate the problem here.
>>
>> matthew black
>> information technology services
>> california state university, long beach
>>
>> From: Ishmael Rufus [mailto:sakamura at gmail.com]
>> Sent: Tuesday, June 26, 2012 9:34 PM
>> To: Matthew Black
>> Cc: David Hubbard; nanog at nanog.org
>> Subject: Re: DNS poisoning at Google?
>>
>> Have you tried using Google Webmaster tools?
>> On Tue, Jun 26, 2012 at 11:28 PM, Matthew Black <Matthew.Black at csulb.edu<mailto:Matthew.Black at csulb.edu>> wrote:
>> Running Apache on three Solaris servers behind a load balancer.
>>
>> I forgot how to lookup our AS number to see if it matches couchtarts.
>>
>> matthew black
>> information technology services
>> california state university, long beach
>>
>> -----Original Message-----
>> From: David Hubbard 
>> [mailto:dhubbard at dino.hostasaurus.com<mailto:dhubbard at dino.hostasaurus
>> .com>]
>> Sent: Tuesday, June 26, 2012 9:14 PM
>> To: nanog at nanog.org<mailto:nanog at nanog.org>
>> Subject: RE: DNS poisoning at Google?
>>
>> Typically if google were pulling your site sometimes from the wrong IP, their safe browsing page should indicate it being on another AS number in addition to the correct one 2152:
>>
>> http://safebrowsing.clients.google.com/safebrowsing/diagnostic?site=ht
>> tp ://www.csulb.edu<http://www.csulb.edu>
>>
>> For example, the couchtarts site they claim yours is redirecting to:
>>
>> http://safebrowsing.clients.google.com/safebrowsing/diagnostic?site=ht
>> tp ://www.couchtarts.com<http://www.couchtarts.com>
>>
>> That site's DNS is screwed up and some requests are sent to a different IP at a different host, so Google picked up both AS numbers.
>>
>> Could one of your domain's subdomains be what is actually infected?  You seem to have a bunch of them, maybe google is penalizing the whole domain over a subdomain?  Not sure if they do that or not.
>>
>> If your sites are running off of an application like wordpress, etc., you may not get the same page that google gets and the application may have been hacked.
>> Here's a wget command you can use to make requests to your site pretending to be google:
>>
>> wget -c \
>> --user-agent="Mozilla/5.0 (compatible; Googlebot/2.1;
>> +http://www.google.com/bot.html)" \
>> --output-document=googlebot.html 'http://www.csulb.edu'
>>
>> nanog will probably line wrap that user agent line making it not correct so you'll have to put it back together correctly.  It will save the output to a file named googlebot.html you can look at to see if anything weird ends up being served.
>>
>> David
>>
>>
>>> -----Original Message-----
>>> From: Matthew Black 
>>> [mailto:Matthew.Black at csulb.edu<mailto:Matthew.Black at csulb.edu>]
>>> Sent: Tuesday, June 26, 2012 11:53 PM
>>> To: nanog at nanog.org<mailto:nanog at nanog.org>
>>> Subject: DNS poisoning at Google?
>>>
>>> Google Safe Browsing and Firefox have marked our website as 
>>> containing malware. They claim our home page returns no results, but 
>>> redirects users to another compromised website couchtarts.com<http://couchtarts.com>.
>>>
>>> We have thoroughly examined our root .htaccess and httpd.conf files 
>>> and are not redirecting to the problem target site. No recent changes 
>>> either.
>>>
>>> We ran some NSLOOKUPs against various public DNS servers and 
>>> intermittently get results that are NOT our servers.
>>>
>>> We believe the DNS servers used by Google's crawler have been 
>>> poisoned.
>>>
>>> Can anyone shed some light on this?
>>>
>>> matthew black
>>> information technology services
>>> california state university, long beach 
>>> www.csulb.edu<http://www.csulb.edu><http://www.csulb.edu>
>>>
>>>
>>>
>>
>>
>>
>
>
>






More information about the NANOG mailing list