Great Circle Associates Firewalls
(June 1994)
 

Indexed By Date: [Previous] [Next] Indexed By Thread: [Previous] [Next]

Subject: RE: Re[2]: SMTP mail spoofing
From: dsmith @ isc . nva . ge . com
Date: Mon, 27 Jun 94 20:56:54 PDT
To: Dave Crocker <dcrocker @ mordor . stanford . edu>, Firewalls @ greatcircle . com

>At 8:13 AM 6/27/94, Dorian W Smith wrote:
>>     If you really need to know about your mail delivery you need X.400.
>>     X.400 and parts of X.500 (either implemented as X.500 or integrated
>
>This is wrong.

Not according to what you write below -- the capabilities are similar.  The difference is
the RSA that is used in X.400/X.500 (X.500 primarily).
>
>It is a common error, but it is still an error.  If the mail is traversing
>uncontrolled paths, I believe that X.400 offers nothing different from what
>is offered by SMTP.

Here you are talking about apples and something else.  What a mail system offers
depends on what you ask it to do.  The X.400 suite can provide a full end-to-end authentication 
and validation.  This includes mail security.  The details are in the '88 and '92 X.500
books.
>
>To the extent that you are referring to X.509-based stuff, then we can also
>cite PEM/PGP/RIPEM/... for Internet (SMTP/RFC822/MIME)-based authentication
>and data integrity.
>
The idea here was not to show superiority of one mail system over another.   Especially
with an Internet group that has grown up with SMTP as a religion.  However, at great
peril, I suggest that security is still X.400 mail systems strong suite, even after SMTP
camps copied some of the stuff out of X.400 (not a bad thing to do when something works).

>In other words, technical solutions exist for both camps.  They just aren't
>being used very heavily.

We use a lot of the X.400 security for our customers.   We delivery turnkey mail systems
among other security related things.
>
>The question is not about the ability to record audit trails, modification
>actions, or the like.  It is about the ability to detect the occurrence of
>these things, independent of the cooperation of the "attacker".
>
This is a multiple layer action.  You really want to catch the would be hacker and the accidental
oops before problem occur.  This means that you must look at all places where attacks can occur
and at all layers.  Check packets to mail-grams and authenticate them all.  We typically have a 
third party (third party in terms of the mail interactions) that watches the transactions.  It records
salient details unless it note something "unusual" and then it kicks into high gear and records 
everything related -- and "rings the crash alarm" to bring attention to the fact that something
is going on.  You as the mail guru have to set the level of false accepts and false detects to the
level your mail customer needs to feel secure.

>Dave
>
>
>Dave
>
>+1 408 246 8253  (fax:  +1 408 249 6205)
>
>
>

-------------------------------------
Name: Dorian W Smith
E-mail: dsmith @
 isc .
 nva .
 ge .
 com (Dorian W Smith)
Date: 27 JUN 94
Time: 2209

This message was sent by Chameleon 
-------------------------------------



Indexed By Date Previous: RE: Did the charter change while I wasn't looking?
From: dsmith @ isc . nva . ge . com
Next: Re: Re[2]: SMTP mail spoofing
From: dsmith @ isc . nva . ge . com
Indexed By Thread Previous: Re: Re[2]: SMTP mail spoofing
From: Frederick M Avolio <avolio @ tis . com>
Next: Re: Re[2]: SMTP mail spoofing
From: dsmith @ isc . nva . ge . com

Google
 
Search Internet Search www.greatcircle.com