[Snort-users] Heartbleed Rule

[Nicholas Bogart]

Boss asked me about creating a rule for the OpenSSL Heartbleed.  I asked him why not just go update all the servers.  He just stared at me.  So I am submitting to the community for review and comment the rule I drew up on this proof-of-concept exploit for the heartbleed vulnerability.

Exploit - https://gist.github.com/takeshixx/10107280

CVE - https://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2014-0160

Heartbleed References -
http://threatpost.com/seriousness-of-openssl-heartbleed-bug-sets-in/105309
https://threatpost.com/openssl-fixes-tls-vulnerability/105300

alert tcp any any -> $HOME_NET 443 (msg:"Attempted Heartbleed access exploitation for OpenSSL 1.0.1f and lower"; flow: to_server; content:"| 18 03 02 00 03 01 40 00 |"; reference:cve, CVE-2014-0160;)

NickyB

[Nicholas Bogart]

Awesome thanks... I tried searching through all my emails for the list to see if this was done.  Glad to see I was on the right track.  Will let the boss know.



Nick

On Wed, Apr 9, 2014 at 1:55 PM, Joel Esler (jesler) <jes...@cisco.com> wrote:

Nick,

Might want to review the latest post on http://vrt-blog.snort.org. 

--

Joel Esler

Sent from my iPhone

------------------------------------------------------------------------------
Put Bad Developers to Shame
Dominate Development with Jenkins Continuous Integration
Continuously Automate Build, Test & Deployment
Start a new project now. Try Jenkins in the cloud.
http://p.sf.net/sfu/13600_Cloudbees

_______________________________________________
Snort-users mailing list
Snort...@lists.sourceforge.net
Go to this URL to change user options or unsubscribe:
https://lists.sourceforge.net/lists/listinfo/snort-users
Snort-users list archive:
http://sourceforge.net/mailarchive/forum.php?forum_name=snort-users

Please visit http://blog.snort.org to stay current on all the latest Snort news!

[Jefferson, Shawn]

Any reason these rules are $EXTERNAL_NET -> $HOME_NET ?  Lot’s of false positives otherwise, performance, or something else? 

 

I was hoping to use them to detect potential internal network heartbleed attacks, but would have to re-write them to do that (never ideal).

 

Thanks

Shawn

[Joel Esler (jesler)]

Not all of them are.  We have rules for both directions.

[JJC]

Beyond what Joel just responded with, if you are looking for internal-internal attacks often you will want your $EXTERNAL_NET variable defined as 'any'.  This would then make the rule direction that you noted effective even for inside -> inside traffic inspection.

JJC

[Jefferson, Shawn]

Thanks everyone!  Makes sense….