[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Future of kerberised telnet, login, rsh, ftp?
>We have similar issues, which is why we support both gssapi-ssh and krcp.
>But one of our network engineers found that most of the slowness in SSH is
>due to poor buffer sizes, not encryption (unless you have a *really* slow
>machine). See:
> <http://www.psc.edu/networking/projects/hpn-ssh/>
Interesting. I'm wondering what your network link is for the second
graph, though, because clearly 23 MB/sec (hpn-ssh with arcfour) is too fast
for 100 Mbps.
If people really want good performance, I tell them to use our version
of kftp, which I added some extensions to allow the user to specify the
send and receive socket buffer size. Using that, I was able to get 50
MB/sec across an OC-12, which is 622 Mbps (I had a couple of terabytes
to move, so I really wanted the best performance I could get).
--Ken
- Prev by Date:
Re: Future of kerberised telnet, login, rsh, ftp?
- Next by Date:
Re: Future of kerberised telnet, login, rsh, ftp?
- Prev by thread:
Re: Future of kerberised telnet, login, rsh, ftp?
- Next by thread:
Re: Future of kerberised telnet, login, rsh, ftp?
- Index(es):