Basicali it is normal for XFS to outperform ext2/3 in your case. High badwidth is the main goal in XFS design ( read about it on sgi.com ). Especialy ext3 in ordered mode. XFS can only work in write-back (the fastest) mode. >From all that I am reading about the linux filesystems, and from my personal experience, it seems that ext2/3 is the slowest comparing to ReiserFS,JFS,XFS. BUT it is the most robust filesystem that you can find for linux. Try this: ------------- mkfs.ext2 /dev/hd?? about 500-1000 MB mount /dev/hd?? /mnt/???? cp -a /usr/doc /mnt/???? or anything else, just fill it with files umount /mnt/???? dd if=/dev/zero of=/dev/hd?? bs=1024 count=100000 fsck.ext2 -f -y /dev/hd?? mount /dev/hd?? /mnt/???? ------------- Now check what is recovered: Allmost everything, but the blocks that was filled with NULLs. Try this with ReiserFS or XFS :)) _______________________________________________ Ext3-users@redhat.com https://listman.redhat.com/mailman/listinfo/ext3-users
