Full Disclosure mailing list archives

Re: Session Sidejacking in facebook


From: "Thor (Hammer of God)" <thor () hammerofgod com>
Date: Sat, 11 Jun 2011 21:47:57 +0000

LMFAO

From: full-disclosure-bounces () lists grok org uk [mailto:full-disclosure-bounces () lists grok org uk] On Behalf Of 
adam
Sent: Saturday, June 11, 2011 1:57 PM
To: Madhur Ahuja
Cc: full-disclosure () lists grok org uk
Subject: Re: [Full-disclosure] Session Sidejacking in facebook

I was actually just kidding about releasing it to the list, but given the nature of the vulnerability - the disclosure 
could have been a lot worse.

"Is this how it works in all social sites ?"

I've personally witnessed countless sites that authenticate a user based on userID/token combination (and nothing 
else). Depending on the actual token length, bruteforcing it is sometimes even possible.

"If the answer is yes, I will be highly doubtful of using internet at
any public place where sniffing or MITM attack is relatively simple to
make."

As you should be, but don't just apply it to social networking sites.

"Are there any measures to prevent it ?"

Servers/applications could do a little more to protect against it (e.g. X token is only valid for Y IP, or by using 
flash cookies as part of the authentication process, etc etc). The difference is, in your example, the IP check 
wouldn't make a difference. Flash cookies aren't necessarily the best route either, for compatibility and other reasons.

On the client side, I'd recommend using a secure VPN connection any time you're accessing the internet from a public 
place/network. You could do that, tunnel over SSH, whatever. The point being: don't send unencrypted data across public 
networks, unless privacy isn't important (e.g. browsing Wikipedia).

On Sat, Jun 11, 2011 at 3:43 PM, Madhur Ahuja <ahuja.madhur () gmail com<mailto:ahuja.madhur () gmail com>> wrote:
Recently, there was a vulnerability discovered in LinkedIn, which is
described here http://www.wtfuzz.com/blogs/linkedin-ssl-cookie-vulnerability/

Basically, this allows someone in network to sniff a cookie value and
apply it in his browses session to hijack the target's user session.

This simple concept even works even in Facebook. I was able to hijack
n number of user's session sitting in my university room in few
minutes.

For every POST request in facebook, similar cookie string is transmitted:

Cookie: datr=09bXXXQ2oOgQuUK0yAzK_JU9; lu=wgj9pmpkAsdXXXTp5vthfh2w;
locale=en_US; L=2; act=13078123502562F3; c_user=xxxxxx;
sct=1123416461; xs=603Afe43db8a71239bd8d7b2a831xxx6241f;
presence=EM307818375L26REp_5f123422481F22X3078XXX1367K1H0V0Z21G307818375PEuoFD769839560FDexpF1307818409174EflF_5b_5dEolF-1CCCC;
e=n

I was able to hijack the remote user's session by just placing the
value of 2 cookies: c_user (which is obviously user id) and xs (seems
like auth token) in my browser session.

Step by step POC:
http://madhur.github.com/blog/2011/06/12/facebooksessionhijacking.html

Cookie: datr=09bXXXQ2oOgQuUK0yAzK_JU9; lu=wgj9pmpkAsdXXXTp5vthfh2w;
locale=en_US; L=2; act=13078123502562F3; c_user=xxxxxx;
sct=1123416461; xs=603Afe43db8a71239bd8d7b2a831xxx6241f;
presence=EM307818375L26REp_5f123422481F22X3078XXX1367K1H0V0Z21G307818375PEuoFD769839560FDexpF1307818409174EflF_5b_5dEolF-1CCCC;
e=n

Is this how it works in all social sites ?

If the answer is yes, I will be highly doubtful of using internet at
any public place where sniffing or MITM attack is relatively simple to
make.

Are there any measures to prevent it ?

Madhur
http://madhur.github.com

_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/

_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/

Current thread: