Hello Stuart ! Thank very much !
With iperf , Now i can see what exactly I would want see I had forgotten the disk write factor. :) I'm doing some tunning's Thank to all that help me Cheers, Guilherme Hakme 2011/3/17 Stuart Henderson <s...@spacehopper.org> > On 2011-03-17, R0me0 *** <knight....@gmail.com> wrote: > > Interesting Rick, I will do tests setting auto on ubiquity, > > It doesn't make a difference to speed whether you use auto or > full-duplex, as long as the ports at both sides of a wire are set > the same way. Generally I recommend leaving things set to auto > (note: don't change duplex settings remotely unless you have a > clear plan of what you're doing, why, and which order to change > things. :-) > > > My doubt ... > > Why windows vista on both sides have speed up to 10mb/s when a do an > > download ? ( I put 2 windows boxes with smb ) > > if I set auto on both sides ( on openbsd boxes ) the speed is 1.8mb/s and > if > > I set 100baseTX ( Half ) the speed is the same > > These are pretty dissimilar tests and it's not clear how you measure > the speed, there could even just be a confusion between Mb/s and MB/s > (1.8MB/s = 14.4Mb/s). > > How about a simpler test where you change fewer variables (and don't > involve the speed of disks on the machines)? Maybe run iperf on both > OpenBSD and Windows, then you're doing the same test on both OS, > and can test OpenBSD->Windows, Windows->OpenBSD, Windows->Windows, > OpenBSD->OpenBSD. > > > > 2011/3/16 Rick Ballard <ideaph...@gmail.com> > >> The Ubiquity is likely an unmanaged device > > oh, they're quite smart actually. *like* :)