logo       

Re: Re: Hardware acceleration using VIA Padlock (was Re: [Openvpn-users] a : msg#00473

network.openvpn.user

Subject: Re: Re: Hardware acceleration using VIA Padlock (was Re: [Openvpn-users] a couple of questions)

"Eric E. Bowles" <eric@xxxxxxxxxx> said:

> Hi James,
>
> > Now on OpenVPN 1.6 or 2.0-beta6 (but not beta5) try a crypto loopback test:
> >
> > openvpn --genkey --secret tmp-key
> > time openvpn --test-crypto --secret tmp-key --verb 0 --tun-mtu 10000
> > --cipher
> > aes-128-cbc
> >
> > This test will generate 10000 random packets starting at a size of 1 byte
> > and
> > going up to 10000 bytes, and loop them back through the the encryption and
> > decryption algorithms.
>
> I got access to a 1GHz VIA C3 box and tried the command using OpenVPN 1.6
> and patched OpenSSL library; here are the results:
>
> Stock Patched
> OpenVPN OpenVPN
>
> real 1m13.084s 0m47.655s
> user 1m12.570s 0m47.560s
> sys 0m0.020s 0m0.010s
>
> I see a 34% improvement with hardware acceleration enabled.

Interesting. Note that this percentage might be off what you will see in real
usage, because --test-crypto is doing other stuff like generating a lot of
strong pseudo-random numbers to fill up the test packets.

OpenVPN is also doing per-packet authentication using the HMAC/SHA1 algorithm,
I wonder if this is hardware accelerated as well?

Can you try a real network test, such as constructing a tunnel between two
locally connected machines, transfering a large file over the tunnel using FTP
(in both directions), and measuring the CPU usage of OpenVPN?

James



-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 -
digital self defense, top technical experts, no vendor pitches,
unmatched networking opportunities. Visit www.blackhat.com


<Prev in Thread] Current Thread [Next in Thread>
Google Custom Search

News | FAQ | advertise