Umount vs. disk spindown

[badhat]

I've recently migrated from a G5 dual to an intel mini. The mini has an external firewire drive (Seagate mechanism) used exclusively for overnight backup using SuperDuper in scheduled mode. The external drive is left unmounted during the day. When the backup occurs, SuperDuper mounts the drive successfully, performs its Smart Copy, and then unmounts the drive.

The problem is that whlle the drive is now unmounted, it does not stop spinning. I have to launch Disk Utility, MOUNT the drive, and then UNMOUNT the drive to cause it to spin down. It remains this way until the next overnight backup, which executes but leaves it spinning.

I'd like to eliminate the spinning of the unmounted drive for a few reasons (noise, wear, etc.), and I'm thinking this could be accomplished with a script. But I'd rather understand more about why it won't spin down when unmounted. Any ideas - anyone?

[dnanian]

Weird. No idea why it would do that!

[badhat]

So, with no additional information, it appears that I should implement a script that first mounts, then unmounts the drive -- and have SD perform this at completion of the backup. Sound right?
If so, suggestions on what that script might look like? Thanks for the (really) rapid response to my initial post.

[dnanian]

But: it already does. It mounts and unmounts: it seems unlikely your script would be any different than ours.

[badhat]

I would agree except for the evidence. I have tried the two scripts (mount and unmount) provided online by Tim Luoma (whom I think you acknowledged).

After SD performs the backup, and the disk remains spinning, I first run the mount script, then the unmount script (both within Terminal) -- the disk then spins down.

So, I remain confounded by this...

[dnanian]

Again, I have no idea why that would be. We're unmounting the volume, with diskutil. It certainly spins down here... perhaps the seagate "power management" software has a weird bug in it?

[badhat]

I wouldn't be at all surprised if the power mgmt software DOES have an effect on this.

I'll just work on the "moun/unmount" script and see if that resolves the issue (redundant as it may seem).

Thanks for the comments.