# What is the space overhead of different filesystems?

I backed up/restored a filesystem from a Ubuntu system to a Fedora 17 one. I saw that according to df result the previously owned room raised by 30 %. What could be the factors for this?

At the Fedora system df display screens: 78 GB made use of

At the Ubuntu system df display screens: 60 GB made use of

Differences in between the systems:

Ubuntu: ext3 (developed years ago)

Fedora 17: ext4 (developed with a vanilla mkfs.ext4 call)

Restoring to an XFS filesystem (on Fedora 17) returns 78 GB made use of room.

Back-up and also recover was performed with GNU tar. The filesystem has a variety of various type of documents (i.e. from resource trees, maildirs to ISO is etc).

9
2022-06-07 14:41:52
Source Share
Answers: 2

The first point that enters your mind is "sparse files". Commonly, one can create a documents with information at a countered right into the documents, and afterwards look for to a much better countered. When creating information to the much better countered, the filesystem would certainly not allocate disk blocks for the interfering offsets. Programs that read those offsets without alloted disk obstructs read absolutely no values.

Tarring up thin documents creates the offsets of a thin documents that does not have actually alloted disk obstructs to allocate disk blocks, both in the tar documents (or result stream) and also in the re - developed documents.

I remember that some DBMS developed thin documents, as did programs like MSC/NASTRAN (limited component modelling system). Supporting those thin documents winds up making use of huge quantities of offline storage space, much to the shock of every person entailed.

10
2022-06-07 15:11:22
Source

Another point to look out for when you see room disparities is room booked for origin (commonly 5% on ext 2,3,4 ). This room permits the OS to function (write log documents, etc ) also if a customer loads the disk (as long as that customer is not origin).

You can watch this setup making use of tune2fs -l:

[[email protected] ]# tune2fs -l /dev/md0 |grep Reserved
Reserved block count:     1279986
Reserved GDT blocks:      1017
Reserved blocks uid:      0 (user root)
Reserved blocks gid:      0 (group root)


You can disable this on your ext filesystems with tune2fs -m 0 /dev/NAME

From checking into it actually promptly, I do not think xfs has actually booked room for origin

3
2022-06-07 15:11:15
Source