If you want to enable ext3 file system support in your custom kernel, you have to choose the following options:
<*> Ext3 journalling file system support [*] JBD (ext3) debugging support
Performance issues
Enabling debugging support implies performance degradation.
To create a new ext3 file system on a newly created partition, just type (as root):
# mke2fs -j /dev/hdx
After the new file system is created you can tune the check intervals:
# tune2fs -i 180 -c 30 /dev/hdx
These are the default values on a Debian system; a check is performed every 180 days or every 30 mounts.
About file system checks
Enabling file system checks while using a journalled file system is not strictly needed, but it's still a safe measure against hardware faults.
Last but at least, an appropriate entry must be added to /etc/fstab.
You can specify different mount options for ext3 file systems.
If you need to convert an existing ext2 file system to ext3, just type (as root):
# tune2fs -j /dev/hdx
After the file system has been converted, you need to change the relative
entry in /etc/fstab, replacing ext2 with ext3 and specifying
the desired mount options.
An ext3 capable kernel will auto detect an ext3 root file system at boot time. If this is unwanted, the following option might be used:
LILO: linux rootfstype=ext2
Mount options for the root file system can be specified with the
rootflags option (see lilo man page for details).
The following options are available for an ext3 file system:
mount -o data=journalmount -o data=writebackmount -o data=orderedThis work is licensed under the terms of the Creative Commons NonCommercial-Attribution-ShareAlike license. A reference copy of this license can be found at the following address: http://creativecommons.org/licenses/by-nc-sa/2.5/