[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: ext3 2.4.21 htree tests


> Why use more directories?  That just means that you spend more time
> updating more metadata for more directories.  The point of htree is to
> let the fs do this work itself, as efficiently as possible: if you try
> with _1_ directory, you should see the best benefit from htree.

Well I the main reason I started spreading files over multiple directories
is from the ext2 days when using a single directory it became unuseable when
it contained more than a couple of hundred thousand files. I've always done
this since and the test results do imply thats its more efficient - thats
why I did the tests with htree as I was curious if this would solve the

> However, the performance still doesn't look that good, especially on the
> 100 dirs case (10,000 files per dir) --- are you sure you set the
> dir-index flag on the filesystem to enable htree operation?

Yes, I'm pretty sure, I did the following:

tune2fs -O dir_index /dev/hda9
e2fsck -fD /dev/hda9

Then queried using dumpe2fs -h /dev/hda9 which gives:

Filesystem volume name:   <none>
Last mounted on:          <not available>
Filesystem UUID:          07691ca4-a168-11d7-8257-00c0df01c990
Filesystem magic number:  0xEF53
Filesystem revision #:    1 (dynamic)
Filesystem features:      has_journal dir_index filetype needs_recovery
Default mount options:    (none)
Filesystem state:         clean
Errors behavior:          Continue
Filesystem OS type:       Linux
Inode count:              17251872
Block count:              17241753
Reserved block count:     862087
Free blocks:              16693285
Free inodes:              17251852
First block:              0
Block size:               4096
Fragment size:            4096
Blocks per group:         32768
Fragments per group:      32768
Inodes per group:         32736
Inode blocks per group:   1023
Last mount time:          Wed Jun 18 16:43:47 2003
Last write time:          Wed Jun 18 16:43:47 2003
Mount count:              3
Maximum mount count:      20
Last checked:             Wed Jun 18 11:14:51 2003
Check interval:           15552000 (6 months)
Next check after:         Mon Dec 15 10:14:51 2003
Reserved blocks uid:      0 (user root)
Reserved blocks gid:      0 (group root)
First inode:              11
Inode size:               128
Journal inode:            8
Default directory hash:   tea
Directory Hash Seed:      7d45fbeb-7e2d-47b2-8ecc-24874e8dc686

If there is anything that I've missed I'd like to know - I'm interested in
getting the best performance possible.

I can run any additional tests.



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]