[lkml]   [2003]   [Aug]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
Subjectide drives performance issues, maybe related with buffer cache.
Hi guys!

Sorry for my post, but i didnt recieve answer on it after some googling &
posting to local Linux Users Group, and i am confident that it is related
with kernel.
I am using linux quite long, tried (&tired of) many FS's on different
hardware, and think this behavior is common.
So question is : why when i am copying file from one HD to another (for
simplicity from /hda to /hdb)
the speed fall down ? Starting from about 27-30 MB/s (drives are in UDMA-4,
hdparm -X68) it drops
down to 11-12 MB/s after 4-5s. In *indows transfer rate is almost constant
and about 20-22 MB/s (same hardware). Why the h#ll we suck?
I feel that it's due to buffer cache, because when you use sync (while
copying) transfer rate is so small or even 0.
Drives are tuned with hdparm to highest transfer rates, readahead, multiple
sector count (hdparm
for details).
Tried different filesystems, from classic ext2/3 to modern xfs/reiserfs. The
same results.
Pure kernel from (2.4.{19,20,21}), vendors kernels - all the
Doing experiments with 'sysctl -a |grep vm' values didnt resolve the
This behavior is general, i think.
And, damn, this is very annoying for end users, for example when copying
large files (movies,iso's).
Beyond any doubt this is very bad for IDE based file servers.

So please help me, and sorry if this post is offtopic.

Linux needs grouppies !
(inspired by Almost Famous & recent akpm interview)

Help STOP SPAM with the new MSN 8 and get 2 months FREE*

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
More majordomo info at
Please read the FAQ at

 \ /
  Last update: 2005-03-22 13:47    [W:0.042 / U:7.088 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site