RE: Very slow scp performance comparing to Linux [dd to /dev/null shows substantial FreeBSD vs. Ubuntu differences for bs=1k (or 1K) and bs=512]

> -----Original Message-----
> From: Mark Millard <marklmi@yahoo.com>
> Sent: Thursday, August 31, 2023 2:17 PM
> To: Mark Saad <nonesuch@longcount.org>
> Cc: Wei Hu <weh@microsoft.com>; FreeBSD Hackers <freebsd-
> hackers@freebsd.org>
> Subject: Re: Very slow scp performance comparing to Linux [dd to /dev/null
> shows substantial FreeBSD vs. Ubuntu differences for bs=1k (or 1K) and
> bs=512]
> 
> On Aug 30, 2023, at 20:59, Wei Hu <weh@microsoft.com> wrote:
> 
> >> -----Original Message-----
> >> From: Mark Millard <marklmi@yahoo.com>
> >> Sent: Thursday, August 31, 2023 10:18 AM
> >> To: Mark Saad <nonesuch@longcount.org>
> >> Cc: Wei Hu <weh@microsoft.com>; FreeBSD Hackers <freebsd-
> >> hackers@freebsd.org>
> >> Subject: Re: Very slow scp performance comparing to Linux [dd to
> >> /dev/null shows substantial FreeBSD vs. Ubuntu differences for bs=1k
> >> (or 1K) and bs=512]
> >>
> >> On Aug 30, 2023, at 18:45, Mark Saad <nonesuch@longcount.org> wrote:
> >>
> >>> All
> >>> Why not take scp out of the picture and try iperf? Why , we could be
> >>> looking
> >> at rss by default in Linux .
> >
> > Actually I did the iperf3 test as well and posted results a couple days ago.
> > Pasting here:
> >
> > FreeBSD iperf3 to localhost, single stream: 30.9 Gb/s Linux iperf3 to
> > localhost, single stream: 48.8 Gb/s
> 
> I'll note that (48.8Gb/s) / (30.9Gb/s) is a much smaller ratio than then either
> of the originally reported Ubuntu/FreeBSD ratios:
> 
> (550 MB/s) / (70 MB/s)
> or:
> (350 MB/s) / (70 MB/s)
> 
> Suggesting one or more non-network issues are contributing, especially vs. the
> localhost test that had the 550/70 ratio.
> 
Yes, the network perf gap is not as big as I originally thought. Seems different
factors contribute to the slowness of scp on FreeBSD. Network stack is only one
of them.

Wei