Cron automatiically backs up some partitions on my HD by means of a script. Not sure of the size of thse backupos, but perhaps 300 Mb.
I have been doing the backups to an external WD USB drive, and they took around 3 hours. However, I became nervous about the condition of the drive which is quite old, and so bought a 2 Tb replacement. Now the back up takes 10 hours. The only thing that I can think of that might account for its being slow is that my old WD drive was formatted ext4, but I thought best to leave my new drive with NTFS. This causes a problem in that if the backup drive happens to be mounted, the mount command in my script no longer just tells me so and proceeds with the backup, but instead hangs. The other problem may be that for some reason the disk being NTFS drastically slows the backup. So it occurred to me to make the command in the script to mount the drive: mount -t ntfs /mnt/backup (I have the drive's UUID in fstab). But when I check /proc/filesystems, ntfs apparently is not recognized by the kernel. However, my impression is that my having the ntfs-3g rw driver installed should enable me to mount a NTFS partion wtihout problem or need for the -t ntfs option. I checked my CPU instuctions/second. The services started at bootime have not changed. The # top command does not show any problems. $ free suggests I'm not demanding too much of my RAM. # iotop shows that my backup process I/O demand on the kernel runs 50-100%. The kworker flush can be 100%. My guess is that these figures are to be expected. I run the backup at a time when no other significant processes are running. -- Haines Brown _______________________________________________ Dng mailing list Dng@lists.dyne.org https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng