From: Marc Maiffret (mmaiffret@eeye.com)
Date: Tue May 04 2004 - 04:07:54 EDT
Okena "works" because no one knowledgeable has said otherwise. Okena has
taken the same flawed approach as network IDS systems focusing on
protecting from exploits, and not vulnerabilities specifically. Although
they have done so by doing detection of exploits at the kernel level,
instead of at the network level.
Take the Microsoft RPC attacks as an example... Okena in its
"behavioral" protection is for the most part protecting against the
recent MS RPC attacks by denying bad "behavior" that is typically seen
within most exploits. Specifically speaking, one of the ways they
"protect" from the RPC attacks is to make sure that calls to
LoadLibrary/GetProcAddy etc.. Are really coming from app code and not
some random place on the heap (for example, could be other places,
obviously...). Yes, this protection works, for now... Because everyone
uses the same templates for exploiting windows flaws that have been used
for years now. The main problem with something like Okena (and most
other kernel-only) systems is that they do not start protecting
applications until AFTER an attacker is executing code... And at that
point it is game over. You can bypass ANY of these kernel protection
systems. You could even use a local windows kernel flaw to do it in some
cases: http://www.eeye.com/html/Research/Advisories/AD20040413D.html
There are many other things wrong with the Okenas of the world, such as
the usability nightmare of most of these "learning" systems. Also the
fact that most of the time they stop an exploit by killing a thread, a
process, or in some cases (Okena has the option) restarting the entire
system. Therefore your "security" is now a system that denies code
execution, in exchange for a denial of service. Which really is not that
much better. I could go on and on but don't have the time now.
And to clarify: I do not think all kernel protection systems are bad,
including Okena. However, kernel protection alone is not enough and only
has a very limited use in what it can do to truly help secure a system,
marketing diagrams aside.
Signed,
Marc Maiffret
Co-Founder/Chief Hacking Officer
eEye Digital Security
T.949.349.9062
F.949.349.9538
http://eEye.com/Retina - Network Security Scanner
http://eEye.com/Iris - Network Traffic Analyzer
http://eEye.com/SecureIIS - Stop known and unknown IIS vulnerabilities
-----Original Message-----
From: Steve Goldsby (ICS) [mailto:sgoldsby@networkarmor.com]
Sent: Monday, April 26, 2004 9:02 AM
To: Steve Goldsby (ICS); Rainer Duffner; Doty, Stephen (BearingPoint)
Cc: pen-test@securityfocus.com
Subject: RE: Why eEye Retina (was MBSA scanner)
CA's eTrust Vulnerability Manager is not a good product.
We had a hard-sell demo in our office, and we were not impressed.
Typical CA sales tactic, you can demo the box, but only for a week, and
only if their engineer babysits it the whole time. It DOES however,
enforce strong policy definition and management, which is where most
organizations fall down. If you don't have policy, this box has less
use.
<get on high horse>
Most organizations (in my experience) simply buy a vulnerability
scanner, run it periodically, and patch what it tells them to patch.
When a patch/fix breaks an application, they back it out. There is
usually very little regard to what other security controls are in place
to mitigate the risk.
Basically, you're playing catchup all the time. Chase the patch, chase
the vulnerability.
</get on high horse>
As an aside: to get around the "chase the patch" mess, look at Cisco
Security Agent (formerly Okena). We run this on all our assets, and we
are running the same binaries and same policy that we loaded **9 months
ago** and we have not had a "hack" yet. No updates, no patches, no
policy changes. We have clients that litteral have a 4 hour maintenance
window each *quarter* and they cannot patch their boxes as patches
become available. Okena/CSA gets around this problem beautifully. And,
it plain works.
This is what personal firewalls should have been doing all along.
Steve Goldsby
www.networkarmor.com
-----Original Message-----
From: Rainer Duffner [mailto:rainer@ultra-secure.de]
Sent: Friday, April 23, 2004 5:50 PM
To: Doty, Stephen (BearingPoint)
Cc: pen-test@securityfocus.com
Subject: Re: Why eEye Retina (was MBSA scanner)
Doty, Stephen (BearingPoint) wrote:
>How does something like CA's eTrust Vulnerability Manager product
compare -
>so that continual scanning is not required using ISS, Nessus, Retina,
etc ?
>
>
>
How does this thing work then ?
I mean, NeVO uses passive scanning, and Nessus-scanning, but this
"thing" ?
Oh, I see:
"Q: How does eTrust Vulnerability Manager detect vulnerabilities? "
"A: eTrust Vulnerability Manager uses non-intrusive methods to detect
vulnerabilities on an asset through a two-step process. Step one is the
identification of technologies running on an asset. This may be
accomplished through manual input or automatically by eTrust"
Vulnerability Manager Service, which identifies the version, patch and
hot fix level of technologies running on an asset. This information is
then correlated with CA s security database to identify the
vulnerabilities that apply to the asset."
Can anyone, who runs this, comment on wether this leads to lots of false
positives/false negatives ?
Does it need an agent ?
And, to be honest, I can't stand "appliances" with specs like that:
"eTrust Vulnerability Manager is an appliance-based solution that runs
on Windows 2000 Server Platform and can be accessed by Internet Explorer
5.0 and higher. "
A 'security-appliance' with the most bug-ridden, most-exploited OS on
the planet, to be used with the most bug-ridden, most-exploited
application running on top of it ?
And:
"In addition, eTrust Vulnerability Manager Service supports: " IBM AIX "
HP-UX " Red Hat Linux " Sun Solaris " Windows NT/2000/XP/Server 2003"
Does that mean it only detects vulnerabilities on those OSs ? What about
all the other stuff that floats around ? The printer that
runs some form of embedded Linux with a vulnerable Apache ?
Rainer
------------------------------------------------------------------------
------
Ethical Hacking at the InfoSec Institute. Mention this ad and get $545
off any course! All of our class sizes are guaranteed to be 10 students
or less to facilitate one-on-one interaction with one of our expert
instructors. Attend a course taught by an expert instructor with years
of in-the-field pen testing experience in our state of the art hacking
lab. Master the skills of an Ethical Hacker to better assess the
security of your organization. Visit us at:
http://www.infosecinstitute.com/courses/ethical_hacking_training.html
------------------------------------------------------------------------
-------
------------------------------------------------------------------------
------
Ethical Hacking at the InfoSec Institute. Mention this ad and get $545
off any course! All of our class sizes are guaranteed to be 10 students
or less to facilitate one-on-one interaction with one of our expert
instructors. Attend a course taught by an expert instructor with years
of in-the-field pen testing experience in our state of the art hacking
lab. Master the skills of an Ethical Hacker to better assess the
security of your organization. Visit us at:
http://www.infosecinstitute.com/courses/ethical_hacking_training.html
------------------------------------------------------------------------
-------
------------------------------------------------------------------------------
Ethical Hacking at the InfoSec Institute. Mention this ad and get $545 off
any course! All of our class sizes are guaranteed to be 10 students or less
to facilitate one-on-one interaction with one of our expert instructors.
Attend a course taught by an expert instructor with years of in-the-field
pen testing experience in our state of the art hacking lab. Master the skills
of an Ethical Hacker to better assess the security of your organization.
Visit us at:
http://www.infosecinstitute.com/courses/ethical_hacking_training.html
-------------------------------------------------------------------------------
This archive was generated by hypermail 2.1.7 : Sat Apr 12 2008 - 10:53:53 EDT