[14824] in bugtraq
Re: glibc resolver weakness
daemon@ATHENA.MIT.EDU (D. J. Bernstein)
Sun May 7 15:36:45 2000
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Message-Id: <20000507021743.32453.qmail@cr.yp.to>
Date: Sun, 7 May 2000 02:17:43 -0000
Reply-To: "D. J. Bernstein" <djb@CR.YP.TO>
From: "D. J. Bernstein" <djb@CR.YP.TO>
X-To: bugtraq@securityfocus.com
To: BUGTRAQ@SECURITYFOCUS.COM
Steven M. Bellovin writes:
[ random ID to make blind DNS packet forgery more difficult ]
> 16 bits was far too small to do it right,
Unpredictable IDs and port numbers make large-scale blind forgeries
vastly more expensive. That's more than DNSSEC has ever accomplished.
See http://cr.yp.to/dnscache/forgery.html for further comments.
> http://www.research.att.com/~smb/papers/dnshack.ps
Cache poisoning is a solved problem. A modern DNS cache simply discards
records outside the server's bailiwick.
---Dan
