filesystem fragmentation stats?

Damian Menscher menscher at uiuc.edu
Sat Jul 9 07:15:41 UTC 2005


On Sat, 9 Jul 2005, evilninja wrote:
> Damian Menscher schrieb:
>> That said, is there a way to check the overall level of fragmentation of
>> a live ext3 filesystem?  I know about filefrag, but that's for specific
>> files.  And I think e2fsck tells you, but only if you take the
>> filesystem offline for the scan.
>
> "tune2fs -l" tells you about "Fragments per group", and "fsck.ext2 -nv"
> opens the fs read-only and print some nice stats after that.

I noticed the "Fragments per group", but haven't been able to find 
anything that documents what it means.  Could someone here comment?

Running e2fsck -nvf gave the info I was looking for:

On my mail partition:
       93 inodes used (0%)
       30 non-contiguous inodes (32.3%)
          # of inodes with ind/dind/tind blocks: 44/17/0
    60540 blocks used (45%)
        0 bad blocks
        0 large files

       80 regular files
        4 directories
--------
       84 files

On my home partition:
   191326 inodes used (7%)
    11084 non-contiguous inodes (5.8%)
          # of inodes with ind/dind/tind blocks: 21649/707/0
  4581594 blocks used (86%)
        0 bad blocks
        0 large files

   174096 regular files
    15787 directories
        5 fifos
      274 links
     1421 symbolic links (1356 fast symbolic links)
        8 sockets
--------
   191591 files


So, interestingly, the home directories haven't gotten too fragmented 
despite being at >90% usage for several months (much of that time at 
>95%).  Apparently the 5% reserved for the system factors in here. 
That's certainly a relief!

On the other hand, the mail spools are getting horribly fragmented 
(>30%), probably because mail programs are deleting messages out of the 
middle of the spools?  It's hard to imagine any other reason for 30% 
fragmentation on a filesystem that's less than half full.  (Another 
system I manage also shows high fragmentation for the mail spool, so I 
think this must be a generic problem.)

Damian Menscher
-- 
-=#| Physics Grad Student & SysAdmin @ U Illinois Urbana-Champaign |#=-
-=#| 488 LLP, 1110 W. Green St, Urbana, IL 61801 Ofc:(217)333-0038 |#=-
-=#| 4602 Beckman, VMIL/MS, Imaging Technology Group:(217)244-3074 |#=-
-=#| <menscher at uiuc.edu> www.uiuc.edu/~menscher/ Fax:(217)333-9819 |#=-
-=#| The above opinions are not necessarily those of my employers. |#=-




More information about the Ext3-users mailing list