Well, I'm the developer of this patch and I fully respect their decision
not to incorporate at this time.
The OpenSSH dev group has, I believe, different priorities. Security is,
and must be, foremost in their mind. Performance rightly takes a back
seat to security concerns. On the other hand, as a developer for high
performance networks I have a client base that needs the performance so
that is foremost in my mind. I don't believe my patch compromises
security but the development group has to be sure of this. Their
reputation will be far more affected than mine if there turns out to be
a problem.
Eventually, I believe that some version of the buffer tuning concept
will be incorporated but it has to meet the high quality requirements of
the development group. In the mean time, early adopters and other people
willing to incorporate this patch are perfectly free to do so. In fact,
even when communicating with non-HPN hosts a performance boost will be
seen if the bulk data flow is in the direction of the HPN patch host. So
even without incorporation of the patch into the main source tree people
can see these advantages.
Development is on going and new version of the patch will be available
shortly. This new version of the patch will hopefully address a
performance issue in LAN transfers.
Chris Rapier
fast wrote:> Hi,
>
> http://www.psc.edu/networking/projects/hpn-ssh/
>
> Clearly, the HPN patches significantly boost throughput performance.
> This enhancement is entirely from tuning the SSH buffer sizes.
>
> Alex Tavcar
>
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> openssh-unix-dev mailing list
> openssh-unix-dev at mindrot.org
> http://www.mindrot.org/mailman/listinfo/openssh-unix-dev