home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!usc!not-for-mail
- From: merlin@neuro.usc.edu (merlin)
- Newsgroups: alt.security
- Subject: Re: Lan watchers and sniffers
- Date: 13 Nov 1992 02:30:55 -0800
- Organization: University of Southern California, Los Angeles, CA
- Lines: 73
- Distribution: alt
- Message-ID: <1e004vINNbtl@neuro.usc.edu>
- References: <BxHts7.3s2@minerva1.bull.it> <1dtgfcINN28a@neuro.usc.edu> <1992Nov12.200500.3166@nic.csu.net>
- NNTP-Posting-Host: neuro.usc.edu
-
- >How would you like if the state was to issue a law tomorrow that
- >exceeeding speed limit results in capital punishment? That's about
- >what you suggest. I would disable an account for a couple of weeks --
- >unless a person has really damaged important files, etc.
-
- In the first place, with promiscuous mode PC based ethernet boards, the
- fact is that there is no absolute physical control over ether snooping.
- There is simply no account to disable in such cases. Moreover, in some
- few environments where password snooping might be controlled by simply
- shutting down some individual user's access code, such minor treatment
- has proven generally ineffective in controlling snooping.
-
- >It's a frustration to see how sites are adopting policies just design
- >to say that users have no rights and {college, university, company}
- >has all priviliges to erase there files, read their love letters and
- >kick them out of {} if they don't like their face. Rather than to
- >reflect real life.
-
- At USC it is my understanding our official policy is to pursue systems
- administrative staff who snoop through user files and personal data as
- virgorously if not more so than we pursue outside penetrators. Our VP
- for computing considers inside or ourside snooping a capital offense--
- you get fired || suspended || excommunicated.
-
- >> ... put known snoopers on their own isolated leg.
- >Aren't they supposed to be shot by now :). Seriously, this is likely
- >to take a lot of time and money - potentially much more that to deal
- >with security break-in.
-
- Depends on what your threshold for damage from penetration may be. If
- you deal in private medical records, engineering design systems which
- may be subject to industrial espionage, and/or defense || intelligence
- data || sources || methods -- then you are going to be more sensitive
- to the hazards of penetration attempts than if you run a public BBS.
- In the case of our university, with sensitive personal information in
- online email records and occasional accidental || determined efforts
- to either steal other student's projects || disrupt computing facility
- so no one can complete their homework on time we get very sensitive.
-
- >I don't remember what original article said, but I recall that software
- >used for debugging was PC based.
-
- If you recall this detail then what's your point about shutting down the
- perpetrator's computer account? Most PC's run DOS based snooper codes
- which require no individual access authorization.
-
- >Anyway, fascist measures above will hardly prevent a person interested
- >enough to do snooping anyway and will certainly make bull.it users
- >hate Alessandro Bottonelli. It looks for me is the only way to avoid
- >password snooping but still permit network debugging is end-to-end
- >encryption. If network is using TCP/IP, original poster may want to
- >look at latest version of BSD telnet[d] ( @gatekeeper.dec.com ) that
- >has built-in encryption.
-
- I hate fascist rules. However, in the absence of adequate funding to do
- a better job of securing our campus systems against intentional and / or
- unintentional penetration/disruption we have to rely on administrative
- controls (what you call fascist rules) and the integrity of our people.
- We cannot afford effective end to end encryption today. Believe me, I've
- done everything possible in our environment to try to adopt tecxhniques
- which would better secure our ethernet without relying on administrative
- controls -- but the only response I get from senior campus administration
- is that we can't afford the cost of implementing these procedures today.
-
- ------------------------------------------------------------------------------
- Alexander-James Annala
- Principal Investigator
- Neuroscience Image Analysis Network
- HEDCO Neuroscience Building, Fifth Floor
- University of Southern California
- University Park
- Los Angeles, CA 90089-2520
- ------------------------------------------------------------------------------
-