On Wed, 30 Oct 2002, Hesham Soliman (EAB) wrote:
>   > > >.. thus making the argument about the ease of use pretty 
>   > much irrelevant
>   > > >IMO ..
>   > > 
>   > > Exactly.
>   > > 
>   > > It makes any argument that site-local filters are more "secure"
>   > > than global filters pretty much irrelevant, too...
>   > > 
>   > > If you can compromise the edge router and change its 
>   > configuration,
>   > > you can get either intra-site global or site-local traffic to be
>   > > forwarded outside of the site.
>   > 
>   > Totally agree; but I'd also add a simpler case: someone forgot to
>   > explicitly configure (or like I did, when reading the spec 
>   > -- assumed that
>   > it should get done automatically) the site scope in the 
>   > edge router(s).
>   > Whoops!  
>   > 
>   > Watching the amount of spoofed traffic nowadays, most of 
>   > which could be
>   > prevented by proper filtering, doesn't give me any reassuration that
>   > people would actually do this too.. and then wonder why 
>   > their private
>   > site-local address space has been compromised..
> 
> => Are you saying that site-local traffic would start
> leaking outside the site and routed globally? 
> As in transient ISPs will just forward it? 

Of course the ISP's will forward them -- they (probably) haven't been
configured to be part of any sites (remember the two interpretations of
the MUST NOT forward paragraph earlier).

-- 
Pekka Savola                 "Tell me of difficulties surmounted,
Netcore Oy                   not those you stumble over and fall"
Systems. Networks. Security.  -- Robert Jordan: A Crown of Swords

--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to