Re: [PEN-TEST] How to "break into" the Pen-Testing field

["Rossman, Hart M." <HART.M.ROSSMAN () SAIC COM>]

I've been using VMWare 2.0 for NT with both Redhat 6.2 and OpenBSD 2.6 for a
couple of months now on a laptop with 128MB Ram. I have the network setup in
the bridged config. I've found it incredibly useful for validating tool
output and obviously using unixes (unices?) robustness without having to
constantly reboot (or schedule/order) the testing steps too carefully.  It
can bog down your NT system though if you don't have enough memory.  One
thing I have had some trouble with though is using sniffers (mostly snort
and NAI Sniffer Pro).  I've had some difficulty validating the accuracy of
the data from the shim drivers.  I haven't tried using the localnetwork
config or a dual/simul network config.

Hart Rossman


-----Original Message-----
From: Teicher, Mark [mailto:mark.teicher () NETWORKICE COM]
Sent: Sunday, September 10, 2000 11:39 PM
To: PEN-TEST () SECURITYFOCUS COM
Subject: Re: How to "break into" the Pen-Testing field


Has anyone considered utilizing a Red Hat Linux 6.2 box running VMWare for
Windows NT.  ???
Instead of having a multi-boot disk or multiple disk packs.

It appears that some of these tools are available for both platforms.

One is absolutely correct, it really depends on the type of penetration
test one is engaged to conduct.  But what really is Industry Best
Practices.  I know some high end consulting services like to utilize a
mixture of commercial and freely available network and host based scanners
to give an overall analysis.  Then parse through the results to formulate a
network and host map.

What tool would be used first and what would be the secondary tool to
validate any false positives one may discover???  Is there any manual
massaging of the data??  Would you turn over the raw data to the customer??

/mark



At 06:17 PM 9/10/00 -0400, Frasnelli, Dan wrote:
> > What would be the typical tool suite one would use on a Pen Test??
>
> I assume you meant the usual network-based penetration test by that.
> If you are asked to mess with a client's pbx/vmb, physical security,
> employees, etc... there are other techniques or hardware involved.
>
> Most penetration tests are conducted in two phases: exploration and
> exploitation.
> I recommend you tailor a software 'tool suite' with those as
> guidelines.  Depending on your style, organizing tools this way may
> or may not be efficient.
> Below are examples biased towards Unix; perhaps an NT person has
> suggestions for that platform.
>
> Exploration and Analysis
> - portscanners:
>         nmap            (www.insecure.org)
> - sniffers:
>         tcpdump         (www.tcpdump.org)
>         ngrep           (sourceforge.net/projects/ngrep)
>         dsniff          (www.monkey.org/~dugsong/dsniff)
> - vuln scanners:
>         vlad            (razor.bindview.com/tools/)
>         whisker         (sourceforge.net/projects/whisker)
> - Samba, nbtscan, l0phtcrack & other tools for windows networks
> - the inevitable custom code and scripts
>
> Exploitation
> - hunt                  (www.gncz.cz/kra/index.html)
> - misc tools            (www.ussrback.com, www.packetfactory.net)
> - whatever is current from packetstorm/ussrlabs/bugtraq/etc.
>   for the targets.
>
> This category is dynamic and typically contains unreleased
> exploits, in-house code, etc.  Its also the attack phase which
> causes most 'script kiddies' grief, as it requires a lot of creative
> tweaking to avoid detection.
>
> A portable computer and disc with various tools compiled for your platform
> of choice is a good starting point for a network penetration kit.
>
> -dan