Security Basics mailing list archives

Re: Internet filtering at the packet level?


From: Rob Creely <programmingart () gmail com>
Date: Sat, 21 Aug 2004 18:50:44 -0400

Any reason you can't just block all outgoing traffic except traffic to
your proxy server which is doing the filtering?

I think you looking at a lot of overhead and network slowdown by
scanning every single packet.  How about the people that connect to a
secure web proxy via SSL?  How about people that use SSH forwarding to
a web proxy?  You can't examine those packets, they are encypted.

Just my 2 cents.....

Cheers.

--Rob 

On Tue, 17 Aug 2004 14:51:25 -0500, Will - Security Engine
<security () the-engine org> wrote:
Ok, I was wondering if it was feasable to filter internet access at the
packet level.  Here is the scenario.

Small college campus - lets say 500 live on campus.  About half that has
internet access.  Then you also have the computer lab, with 16
computers.  Each teacher has a computer in their office as well, and the
CIS dept has about 30 or so computers in use.

The filtering would be done on a Linux server using TCPDump.  I know how
to implement flags for content checking (If the phrase "hot monkey sex"
comes up in a packet, the user is flagged and traffic for that user
would be logged for a set period of time for reviewing later).  What I
don't know is how to actually stop the traffic - but we won't worry
about that for now.

Is there any problems with this?  Is it feasable?  How about just the
flagging portion of it, rather than the actual content blocking?

I'm a student at a private baptist college that gets it's internet
access through MOREnet.  They require that we filter the content in
order to use their services.  Currently we only use a URL keyword and
blacklist filtering system (from my own tests), but it's obvious that
anybody who is serious about getting around the filter will have no
problem (web proxies are stupid easy to set up yourself, and P2P isn't
filtered).  I'm worried that at some point it will come up that we
aren't doing a good enough job filtering, so we'd need a new solution.
I think the packet-based system would be more accurate.  I would be more
inclined to not actually block the content that gets flagged.  I would
rather know that the user is accessing content ruled against by the ToS
and confront them on the issue.

Lets not turn this into a censorship debate please ;)

---------------------------------------------------------------------------
Computer Forensics Training at the InfoSec Institute. 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. 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 so that it never happens again.

http://www.securityfocus.com/sponsor/InfoSecInstitute_security-basics_040817
----------------------------------------------------------------------------



---------------------------------------------------------------------------
Computer Forensics Training at the InfoSec Institute. 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. 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 so that it never happens again.

http://www.securityfocus.com/sponsor/InfoSecInstitute_security-basics_040817
----------------------------------------------------------------------------


Current thread: