1996-06-06 - Re: How can you protect a remailer’s keys?

Header Data

From: lcs Mixmaster Remailer <mix@anon.lcs.mit.edu>mix-admin@anon.lcs.mit.edu
To: cypherpunks@toad.com
Message Hash: 70918dafdbe04b79dcaf5c5cfcef7c152a32645ee47de6b1df546aff4f59cc68
Message ID: <199606051600.MAA02582@anon.lcs.mit.edu>
Reply To: <addac392000210049493@[206.170.115.3]>
UTC Datetime: 1996-06-06 02:21:23 UTC
Raw Date: Thu, 6 Jun 1996 10:21:23 +0800

Raw message

From: lcs Mixmaster Remailer <mix@anon.lcs.mit.edu>mix-admin@anon.lcs.mit.edu
Date: Thu, 6 Jun 1996 10:21:23 +0800
To: cypherpunks@toad.com
Subject: Re: How can you protect a remailer's keys?
In-Reply-To: <addac392000210049493@[206.170.115.3]>
Message-ID: <199606051600.MAA02582@anon.lcs.mit.edu>
MIME-Version: 1.0
Content-Type: text/plain


-----BEGIN PGP SIGNED MESSAGE-----

> From: loki@infonex.com (Lance Cottrell)
> Date: Tue, 4 Jun 1996 22:02:11 -0700
> 
> The best solution I could come up with (and was willing to write and use)
> is to specify the passphrase on the command line argument to the compiler
> 
> make solaris -DPASS="foozooblue"
> 
>                 -Lance

A far better solution would be to have a long-running daemon hold the
secret key.  The mixmaster client could talk to the key daemon through
a unix-domain socket with the permission bits set such that only the
mixmaster user can connect.  Each time the machine is rebooted, the
operator must start the daemon and give it a passphrase.

This has two advantages.  First, it's a lot harder to back up the key
by accident.  If the key ever starts making it only your daily
backups, you are completely hosed because erasing a bunch of mag tapes
would take a lot of time--and maybe you also want to keep your
backups.

Second, if your machine is seized or someone gains unauthorized
physical access to it, the easiest way to get a root shell is by
rebooting single-user.  However, if the only cleartext copy of a key
is in memory rather than in the filesystem, once the machine is
rebooted the secret key is lost.

- - mix-admin@anon.lcs.mit.edu

-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
Comment: Processed by Mailcrypt 3.4, an Emacs/PGP interface

iQCVAwUBMbWrj0TBtHVi58fRAQEkdQP/e7mouEmphgDmn0NKbaCM4lYnT2WbCFsk
irM2GjttiBdpQxr2QDJKBgmHnuGc09xdiexnGnn4bDFie70YDH2Zma3xF/0OvZeQ
DcgAz/0XwkAGPeLCSg8gfeykWwC0HUJlvGtmOwTQKFn5XtlqFM7pKIYF7lnFtoGY
AX/GoGauum4=
=rhyW
-----END PGP SIGNATURE-----





Thread