Bug 53741 - Used block count in df output
Summary: Used block count in df output
Status: CLOSED INVALID
Alias: None
Product: File System
Classification: Unclassified
Component: ext4 (show other bugs)
Hardware: All Linux
: P1 low
Assignee: fs_ext4@kernel-bugs.osdl.org
URL: http://marc.info/?l=linux-ext4&m=1360...
Keywords:
Depends on:
Blocks:
 
Reported: 2013-02-13 05:15 UTC by mujeeb.adil
Modified: 2013-11-18 15:39 UTC (History)
3 users (show)

See Also:
Kernel Version: 2.6.32
Subsystem:
Regression: No
Bisected commit-id:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Description mujeeb.adil 2013-02-13 05:15:23 UTC 
I have an observation on EXT4 filesystem. I created filesystem of size
1TB, 4TB, and 7TB and then checked the output of df command.

df command showed the number of 1KB blocks used. The result was:
1TB: 204056
4TB: 198680
7TB: 181784

I performed the same on XFS and the result was:
1TB: 32928
4TB: 32928
7TB: 33024

EXT4 result shows with increasing filesystem size, the number of used
blocks decreased. This looks odd.

From the linux-ext4 mailing list discussion (http://marc.info/?l=linux-ext4&m=136021915127138&w=2), it seems to be a bug in df reporting.
Comment 1 Eric Sandeen 2013-02-13 05:21:01 UTC 
Specifically, I think this is a problem with the overhead calculations done in the default "bsddf" freespace reporting, which tries to not count the fundamental filesystem metadata blocks as "used".  With mount -o minixdf, which is the straightforward "report all blocks used as used" mode, the df output makes sense.
Note You need to log in before you can comment on or make changes to this bug.