Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.user > #229943

Re: transfer speed data

Show all headers | View raw

[Multipart message — attachments visible in raw view] - view raw

On 23/12/20 11:03 pm, Michael Stone wrote:
> On Wed, Dec 23, 2020 at 09:56:01AM +0800, Jeremy Ardley wrote:
>> Having said that, scp and ssh are affected by the encryption 
>> algorithm. The
>> fastest one at the moment is blowfish and it's possible to get up to 
>> 50 MB/s on
>> a gig lan.
>
> That's pretty ancient advice. The fastest on most modern x86 CPUs with 
> AES-NI instructions is aes128-gcm@openssh.com. Without AES-NI your 
> fastest may be chacha20-poly1305@openssh.com. The default is chacha20, 
> which is fast enough in most cases that it doesn't matter, but worth 
> testing & reconfiguring in cases where it does. Blowfish isn't 
> supported in the latest versions of ssh, and even before it was 
> dropped it was much slower than hardware-accelerated AES. It also 
> never got an authenticated encryption mode IIRC, so it had additional 
> MAC overhead that the more modern modes do not.
>
> The following are on a mid-range Ryzen machine running to localhost, 
> to take the network out of the equation, and are copying a sparse 1G 
> file to /dev/null so there's no disk I/O; either of these algorithms 
> will easily max out a gigabit connection if the disks are fast enough.
>
>> scp -o Ciphers=aes128-gcm@openssh.com testfil localhost:/dev/null
> testfil                               100% 1024MB 864.3MB/s 00:01
>> scp -o Ciphers=chacha20-poly1305@openssh.com testfil localhost:/dev/null
> testfil                               100% 1024MB 475.1MB/s 00:02
>
> For comparison, here's stretch (still supported blowfish) on a much 
> lower power intel CPU (i3-7100U):
>
> $ scp -o Ciphers=chacha20-poly1305@openssh.com testfil 
> localhost:/dev/null
> testfil                               100% 1024MB 167.7MB/s 00:06    $ 
> scp -o Ciphers=aes128-gcm@openssh.com testfil localhost:/dev/null
> testfil                               100% 1024MB 507.5MB/s 00:02    $ 
> scp -o Ciphers=blowfish-cbc testfil localhost:/dev/null
> testfil                               100% 1024MB  77.8MB/s 00:13
> (see how terrible blowfish is, and how the AES-NI acceleration leads 
> to AES tremendously outperforming CHACHA20?)
>
> here's an almost 10 year old non-AES-NI desktop cpu:
>
> $ scp -o Ciphers=aes128-gcm@openssh.com testfil localhost:/dev/null
> testfil                               100% 1024MB 224.7MB/s 00:04    $ 
> scp -o Ciphers=chacha20-poly1305@openssh.com testfil localhost:/dev/null
> testfil                               100% 1024MB 184.9MB/s 00:05
> Note that AES & CHACHA20 are much closer in performance, but AES is 
> still faster. Note also that either can still max out gigabit ethernet.
>

Thanks for the update. Here is my available cipher list:

ssh -Q cipher
3des-cbc
aes128-cbc
aes192-cbc
aes256-cbc
rijndael-cbc@lysator.liu.se
aes128-ctr
aes192-ctr
aes256-ctr
aes128-gcm@openssh.com
aes256-gcm@openssh.com
chacha20-poly1305@openssh.com

I did some tests and found there was around a 10-20% difference in speed 
between runs. This is on a Ryzen 5 with M.2 PCIe drive using 
aes128-gcm@openssh.com

(base) jeremy@client:~$ scp -o Ciphers=aes128-gcm@openssh.com 
sparse_file localhost:/dev/null
sparse_file 100% 1024MB 770.8MB/s   00:01
(base) jeremy@client:~$ scp -o Ciphers=aes128-gcm@openssh.com 
sparse_file localhost:/dev/null
sparse_file 100% 1024MB 814.2MB/s   00:01
(base) jeremy@client:~$ scp -o Ciphers=aes128-gcm@openssh.com 
sparse_file localhost:/dev/null
sparse_file 100% 1024MB 757.6MB/s   00:01


-- 
Jeremy

Back to linux.debian.user | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

transfer speed data mick crane <mick.crane@gmail.com> - 2020-12-23 02:00 +0100
  Re: transfer speed data Bob Weber <bobrweber@gmail.com> - 2020-12-23 02:40 +0100
  Re: transfer speed data Georgi Naplatanov <gosho@oles.biz> - 2020-12-23 02:50 +0100
  Re: transfer speed data Jeremy Ardley <jeremy@ardley.org> - 2020-12-23 02:50 +0100
    Re: transfer speed data Jeremy Ardley <jeremy@ardley.org> - 2020-12-23 03:00 +0100
      Re: transfer speed data Michael Stone <mstone@debian.org> - 2020-12-23 16:10 +0100
        Re: transfer speed data Jeremy Ardley <jeremy@ardley.org> - 2020-12-23 16:40 +0100
          Re: transfer speed data Michael Stone <mstone@debian.org> - 2020-12-23 17:00 +0100
            Re: transfer speed data Jeremy Ardley <jeremy@ardley.org> - 2020-12-23 17:20 +0100
              Re: transfer speed data Michael Stone <mstone@debian.org> - 2020-12-23 18:40 +0100
  Re: transfer speed data Andy Smith <andy@strugglers.net> - 2020-12-23 05:00 +0100
    Re: transfer speed data Nicolas George <george@nsup.org> - 2020-12-23 11:00 +0100
      Re: transfer speed data Andrei POPESCU <andreimpopescu@gmail.com> - 2020-12-23 15:50 +0100
        Re: transfer speed data Nicolas George <george@nsup.org> - 2020-12-23 17:10 +0100
        Re: transfer speed data David Wright <deblis@lionunicorn.co.uk> - 2020-12-23 18:50 +0100
          Re: transfer speed data Andrei POPESCU <andreimpopescu@gmail.com> - 2020-12-23 19:20 +0100
            Re: transfer speed data David Wright <deblis@lionunicorn.co.uk> - 2020-12-24 02:30 +0100
              Re: transfer speed data Andrei POPESCU <andreimpopescu@gmail.com> - 2020-12-24 10:10 +0100
                Re: transfer speed data Michael Stone <mstone@debian.org> - 2020-12-24 17:10 +0100
      Re: transfer speed data Michael Stone <mstone@debian.org> - 2020-12-23 16:20 +0100
        Re: transfer speed data Nicolas George <george@nsup.org> - 2020-12-23 17:20 +0100
          Re: transfer speed data Michael Stone <mstone@debian.org> - 2020-12-24 17:10 +0100
    Re: transfer speed data deloptes <deloptes@gmail.com> - 2020-12-23 12:20 +0100
  Re: transfer speed data George Shuklin <george.shuklin@gmail.com> - 2020-12-23 23:00 +0100

csiph-web