Locking out Bad Tracks

[Dave Hughes]

 Hey, I have just installed SCO Unix 3.2.2 (sys 5) on a 333 mb Maxtor 
hard disk in a 33 mhz 386 and there are a handful of bad sectors on  
the disk all of which have been repeatedly reported by destructive   
low level scanning and automatic entry into the bad track table.     
But upon each boot, after the hardware status is reported, but       
before the SINGLE USER mode prompt comes up, it reports the bad sector
it seems to be trying to read, including an ominous report that      
it had trouble in the swap space reading the 'memsize' value. The    
tracks are well into the swap table, so not right at the beginning.  

After 3 re-installs it does the same thing.                        

Then (2) it has had trouble with bad sectors under the 'goodpw' 
program in bin, which has made some logons impossible. And with
the incredibly complicated UNIX tcb security features, it has    
taken too much time repairing corrupted files caused by reboots
without shutdowns. Here again it attempts to read a 40k utility
on a throughly reported set of bad tracks and bombs over other
time (and prevents even a tape backup past the point).

  Is this bizzare or what? I thot that unix would read the        
badtrack table, alias the sectors to some other tracks and keep   
on truckin.                                                       

 So what does it take to *REALLY* lock out bad tracks?                
(there are really very few reported for the huge drive, only
about 10 seperate entries. And all the problems are on already
tabled badtracks, not on any new ones.)