Text 6973, 35 rader
Skriven 2006-08-10 05:07:12 av Maurice Kinal (1:140/13.1)
   Kommentar till text 6965 av Paul Rogers (1:105/360.0)
Ärende: You gotta smile
=======================
Hey Paul!

Aug 09 18:39 06, Paul Rogers wrote to Maurice Kinal:

 PR> What happens if you diff the disk with /dev/zero again?

It should fail but I didn't bother reading it back since the -cc switch to
mke2fs does that for me.  I did do that with the 64M compactflash disk seeing
it is small and it passed that test with flying colours as well as with the -cc
mke2fs switch.  I am happy with that scheme for smaller disks (just a
write/read dd cycle ... or two or more if really paranoid).

Most of the time (really bad disks) won't get by the dd write, nevermind a
read, but I did find ones that did which were then caught by the -cc switch (-c
switch also caught them on the first read and only does two write/read cycles).
 Some of them are probably salvageable by marking badblocks but to be honest I
have no interest in doing that.  I have enough good drives that it isn't an
issue with me.  The neighbour might try and salvage the more useable ones for
whatever but he doesn't seem overtly keen about it for whatever reason.  I am
betting they end up collecting dust.

Bottomline is it depends on the size and speed of the drive how intensive
testing one cares to get with it but if the first write doesn't catch it the
first read and comparison caught every error.

What do you think?  A pure dd solution?  Offhand I think prefer the one dd
write and then let mke2fs catch anything missed with the -c switch, and -cc if
I really want to make sure.  Eventually it has to get formatted so it seems to
me the intesive tests are better run then.

Life is good,
Maurice

--- Msged/LNX 6.2.0
 * Origin: The Pointy Stick Society XXII - A two bit outfit (1:140/13.1)