Well my iPod coughed and died again.... probably because Freya had been using it as a hammer to bash some toy or other (with it switched on).... I tried everything to revive it...From firing up Linux and using gparted and disk utility, reformatting it with foreign file systems, running badblocks on it, you name it I ran it.... I even loaded DOS USB drivers via the ASPI interface and tried to run Spinrite on it but it still failed....
Then I went into the iPod diagnostics (reset followed by holding down rewind and select whilst it was booting) .... I went into the IO area and looked at the smart data which said it had :-
Retracts = 86
Reallocs = 371
Pending sectors = 151
power on hours = 1127
start stops = 10703
so obviously (to me at least) it knew it had bad blocks but it hadn't remapped them, the difficulty was telling the hard drive where it's bad blocks were .... I have to admit I stumbled across it by accident, you can't "really" access the iPod at such a low level because the operating system "thinks" it's only 7.1 GB so obviously the firmware on the iPod does some fancy remapping of the disk surface via the partition....
Ahhh I thought.... the iPod firmware must look after bad sectors simply by mapping them out at filesystem level not hardware level.... so each time you format your iPod you must lose this bad block map, hence why you always get back to square one each time you restore your iPod... so I experimented and tried to sync the iPod several times.... each time I rebooted the iPod (after it failed) and looked at the diagnostic smart data.... low and behold each time I did this the Reallocs went up and the pending sectors went down....I repeated this until the iPod started syncing without failing....
My iPod is now syncing happily after remapping the bad area(s) on the disk (automatically I might add) ... once it's finished I'll recover the iPod again and see if the bad blocks stay remapped... Well..... I like to prove a theory rather than take it as gospel
No comments:
Post a Comment