The ext3 journal defaults to writing in 'ordered' mode. From the ext3 FAQ:
[Ordered mode o]nly journals metadata changes, but data updates are flushed to disk before any transactions commit. Data writes are not atomic but this mode still guarantees that after a crash, files will never contain stale data blocks from old files.
Changing to 'writeback' mode can improve performance. Again from the ext3 FAQ:
[Writeback mode o]nly journals metadata changes, and data updates are entirely left to the normal "sync" process. After a crash, files will may contain stale data blocks from old files: this mode is exactly equivalent to running ext2 with a very fast fsck on reboot.
As a result, 'writeback' mode will perform better at a cost of potential data loss. If data integrity is critical, do not switch from ordered mode. However, if you have regular backups, and are willing to sacrifice the potential for getting stale data upon reboot after a crash, you can enable writeback mode for performance gains.
Before attempting to enable writeback mode, please be advised performing these operations on a currently mounted hard drive can cause damage and data loss, if you need to perform these operations on the main system drive, you should use a live cd and perform the operations while booted to it (I prefer Knoppix, but you can use whichever you prefer.)
To enable writeback mode type (as root):
tune2fs -o journal_data_writeback /dev/X
where X is the hard drive device and partition in question (hda1, sda1, etc.), then edit /etc/fstab (as root), and under the options section, add "data=writeback" and remount your drive. You can check to make sure that your drive has been mounted in writeback mode by typing:
tune2fs -l /dev/X
as root and looking at the Default Mount Options (it should say journal_data_writeback).
To switch back to ordered mode type (as root):
tune2fs -o journal_data_ordered /dev/X
then remove the "data=writeback" from the appropriate /etc/fstab line.