sjh - mountain biking running linux vegan geek spice - mtb / vegan / running / linux / canberra / cycling / etc

Steven Hanley hackergotchi picture Steven
Hanley
About

email: sjh@svana.org

web: https://svana.org/sjh
twitter: https://twitter.com/sjhmtb
instagram: https://instagram.com/sjhmtb

Other online diaries:

Aaron Broughton,
Andrew Pollock,
Anthony Towns,
Chris Yeoh,
Martijn van Oosterhout,
Michael Davies,
Michael Still,
Tony Breeds,

Links:

Linux Weekly News,
XKCD,
Girl Genius,
Planet Linux Australia,
Bilbys,
CORC,

Canberra Weather: forecast, radar.

Subscribe: rss, rss2.0, atom

←January→
Mon Tue Wed Thu Fri Sat Sun
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 31        

←2007→
Months
JanFeb Mar
Apr May Jun
Jul Aug Sep
Oct Nov Dec

Categories:

Archive by month:

Tue, 09 Jan 2007

The sync option to mount does not mix well with vfat and memory cards - 16:41
So I was wondering why the performance of USB memory sticks appeared to be so pathetic on my laptop and my desktop at work the other day. Read performance was fine with 10 or 1 MB per second, however depending on which memory stick I used I got between 70 KB/s and 600 KB/s.

After banging my head against this for a while I googled for details about bad usb memory performance on Linux. I came across a lkml thread from may 2005 that seems to have helped enough. Apparently the performance of USB memory with the sync option and vfat filesystems is really pathetic, this is largely due to the repeated hammering of 2 blocks with every sync.

Alan Cox has some good and salient points in the discussion (to be expected from such a guru I guess), notably he points out most quality flash memory is very unlikely to be too adversely affected in a short time by using sync and he has a link to some details of life time guarantees from some companies for their flash products.

Anyway I disabled sync on the desktop image and my own desktop and disabled it on my laptop, all of a sudden I get 2MB/s or better depending on the memory stick I am using. Neato.

Interestingly Alan suggests the documentation for mount is generated form the kernel docs somehow and should be up to date and thus not continue to suggest that vfat filesystems ignore the sync flag. It is interesting to see that my Debian unstable copy of that man page on my laptop today still suggests that vfat ignores the sync option. At a glance I can not see any mention of this on the Debian bugs page for mount.

[/comp/hardware] link

home, email, rss, rss2.0, atom