on hard drive failures

[Jim Kuzdrall]

> * Is that part of the hard drive especially weak due to geometry? 
> That would suggest placing superblocks elsewhere.

    Yes, but it doesn't have to be.

    The signal picked up from a magnetic domain is proportional to the 
velocity it moves by the head.  For a platter turning at a constant 
angular velocity, both the length of the track and the velocity are 
proportional to the radial distance of the track from the center.

    In the old days, the disk hubs were large, so the radius varied by 
about 2-to-1.  With the high capacity disks, they probably work much 
closer into the center.

    The old disks used "pre-emphasis" on the inner tracks to make up for 
the speed/density crunch.  It sounds like the drive designers should be 
cutting down further on the sectors/track of the inner tracks.  They 
may have pushed it too far.

    Assuming they start using the disk from the center, a simple way to 
get data at the outer edge might be to take 98% of the disk in the 
working partitions and use the last 2% (presumably the outer tracks) to 
back up the superblocks and other vital information.

   Is that idea worth an experiment?

Jim Kuzdrall