Ethical Hacking Learn to find vulnerabilities before the bad guys do! Gain real world hands on hacking experience in our state of the art hacking lab. Course designed and taught by expert instructors with years of penetration testing experience. 12 student maximum in every class. Certification attempt included in every package. | Computer Forensics Training at InfoSec Institute Gain the in-demand skills of a certified computer examiner, learn to recover trace data left behind by fraud, theft, and cybercrime perpetrators. Discover the source of computer crime and abuse at your organization so that it never happens again. All of our class sizes are guaranteed to be 12 students or less to facilitate one-on-one interaction with one of our expert instructors. |

| Subject: | Re: detecting network crowd surges |
|---|---|
| Date: | Tue, 8 Aug 2006 09:40:54 +1000 |
I'm curious to get some feedback on detecting zombie networks and such by looking at common unique destination IP/port combinations for control and "phone home" traffic.
The idea is to watch a large population of "good guys" like all of the user IPs on an ISP's cable modem network or all of the IPs at a university and detect when ~100 or more all go to IRC, an FTP site, SSH, .etc all in the same time frame.
We've written some correlation rules for our log analysis products to do this in realtime with firewall, network, ids, netflow, .etc traffic, and are getting all sorts of results. I have a blog entry on it (including some screen shots) at:
http://blog.tenablesecurity.com/2006/08/detecting_crowd.html
Sometimes the results are very conclusive, such as ~50 different IPs all checking into IRC at a certain time or all SSHing into an IP address for a second or so.
We've also been able to discriminate this sort of activity on web/ssl traffic by changing some of the thresholds. Occasionally, you can see false positives such as everyone hitting Google or MySpace in a short amount of time. Also, some P2P apps, Skype and others do seem to behave in this sort of 'surge' manner.
Most of the operational stuff I've run across for detecting botnets is either looking at inbound/outbound IDS alerts or running a honeypot. I think those approaches just skim the surface of all the different ways to manage a botnet. A good paper on a broader approach is:
http://www.eecs.umich.edu/~emcooke/pubs/botnets-sruti05.pdf
I'm curious operationally, what other people are detecting. We all run NIDS, SIMS and NBAD products right? What happens to your logs when someone fires up bittorrent, emule, skype, tor, .etc and what happens when you have a real botnet?
I wonder, though, is this how real botnets are controlled?
Surely it would be fair easier, and less obtrusive, to control your botnet via a updated http site. like http://<mikeiscool>/instructions.txt. Every day the bots would log on and receive their latest orders. Makes sense to hide in http rather then risk a protocol that might be blocked, doesn't it?
-- mic
------------------------------------------------------------------------ Test Your IDS
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: anomaly vs signature, Rodrigo Blanco |
|---|---|
| Next by Date: | NNTP and Yahoo IM conflict., NTR |
| Previous by Thread: | detecting network crowd surges, Ron Gula |
| Next by Thread: | Re: detecting network crowd surges, Jose Nazario |
| Indexes: | [Date] [Thread] [Top] [All Lists] |