Computational efficiency is one of the criteria for selecting Rijndael as
AES. Below is the output of openssl speed from my desktop:

The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192
bytes
des cbc          53346.83k    55115.71k    55865.17k    56023.04k
 56027.82k
des ede3         20305.48k    20581.74k    20687.45k    20645.55k
 20679.34k
aes-256 cbc      56244.68k    80301.78k    90277.29k    93125.63k
 94674.94k
aes-128-cbc      83307.89k   114343.85k   126573.48k   130061.99k
130512.21k

If you are referring to the network transfer speed, then CPU time is
probably not the bottleneck:)

On Fri, Jun 26, 2009 at 10:08 PM, George <[email protected]> wrote:

>
> Hi Freepine,
>
> That did the trick, thanks! I was surprised I didn't see much of a
> speed increase in the (non-cached) connection being established when I
> dropped down from a 256-bit AES to a 40-bit export DES key. I guess
> most of the time is gobbled up by the handshake itself.
>
> On Jun 20, 3:01 am, Freepine <[email protected]> wrote:
> > Hi, SHA1PRNG indicates the preudo-random number generator is based on
> > SHA-1(a message digest  algorithm), it's not the encryption algorithm
> used
> > for communication.
> > You can try SSLSocket.getSupportedCipherSuites() to retrieve the
> information
> > of supported cipher suites, and setEnabledCipherSuites to specify the
> cipher
> > suites enabled for use on this connection.
> >
> > On Fri, Jun 19, 2009 at 11:22 PM, george <[email protected]> wrote:
> >
> > > As far as I can see, the only encryption algorithm available for SSL
> > > connections is the default SHA-1. (Which can be passed as
> > > SecureRandom.getInstance("SHA1PRNG") into the SSLSocketFactory
> > > constructor. Is there support for other algorithms such as DES, or
> > > RC4?
>

Reply via email to