No subject

Based on a Pentium 3 800 MHz - 60 GB IDE drive -- UDMA (66)

Crypto                   Write speed      % Loss           Read speed           =
 % Loss
Unencrypted            24619 kB/s       --                  25273 kB/s          =
     --
AES-128bit             20880 kB/s     15.2%               8054 kB/s           =
68.1%
AES-256bit             17699 kB/s     28.1%               8031 kB/s           =
68.2%
3DES-64bit               4846 kB/s     80.3%               3334 kB/s           =
86.8%
3DES-192bit             4869 kB/s     80.2%               3334 kB/s           =
86.8%
BLOWFISH-128bit 19769 kB/s    19.7%               8029 kB/s           68.2%
BLOWFISH-256bit 19755 kB/s    19.8%               8047 kB/s           68.2%

That agrees with your slowdown estimates.

Greg

 >>  On Wed, Mar 12, 2003 at 05:08:49PM -0500, Greg Freemyer wrote:
 >>  > All,
 >>  >=20
 >>  > I know this is off-topic, but does anyone know where I can find info on
 >>  the CPU hit from using a software encrypting FS.

 >>  well, i'll quickly bring in some LVM stuff, as i run encryption both under
 >>  and ontop of a LVM-system. (different cihper though).


 >>  > In particular I would like to see a comparison between None, Microsoft
 >>  EFS, and Linux loopback encryption.

 >>  well, isnt EFS a FS encryption? loopback is a block encryption, there is a
 >>  difference. My experience is that i with loop-AES or PPDD, (blowfish as
 >>  cipher) gets about 1/4 to 1/3 of the performance of raw.



 >>  JonB

 >>  _______________________________________________
 >>  linux-lvm mailing list
 >>  linux-lvm at sistina.com
 >>  http://lists.sistina.com/mailman/listinfo/linux-lvm
 >>  read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/