Date: Mon, 18 May 2020 14:35:12 +0000 From: bugzilla-noreply@freebsd.org To: usb@FreeBSD.org Subject: [Bug 244356] Writing to a USB 3.0 stick is very slow Message-ID: <bug-244356-19105-dc0WwE2eiT@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-244356-19105@https.bugs.freebsd.org/bugzilla/> References: <bug-244356-19105@https.bugs.freebsd.org/bugzilla/>
next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D244356 --- Comment #52 from Olivier Certner <olivier.freebsd@free.fr> --- (In reply to Maurizio from comment #48) Thanks Maurizio. I have roughly the same output as yours with my KT_32G (Kingston DataTraveler 100 G3, showing up as <Kingston DataTraveler 3.0> in kernel logs), *provided* that I set I/O size to anything but "Random" (I ne= ed to investigate this). In this case, the read speed reported by `usbtest` happens to be fairly sta= ble at 110MiB/s, whereas write speed is quite bumpy and ranges from 5 to 40MiB/= s. To me, it seems that your trace essentially shows the same performance prof= ile. The write speed in particular is not a lightning one, but could be consider= ed acceptable. What is very strange, however, is that an `iostat` recording of a `dd` to e= xFAT shows that transfer again proceeds at decent speeds during ~2s, and then st= alls for ~2s, and the pattern goes on and on. And I observe this on Linux and Ma= cOSX as well. Hans, do you consider this pattern as normal? I never observed such one with USB 2.0 sticks so far. Maurizio, if you have time, could you try the `dd` (with, e.g., parameters `if=3D/dev/zero of=3D<file_on_stick> bs=3D1M count=3D1024`) test above and = capture the output of `iostat -w 1 <stick_device>` during the test? --=20 You are receiving this mail because: You are the assignee for the bug.=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-244356-19105-dc0WwE2eiT>