FC4 kernel performance

[Rodd Clarkson]

On Thu, 2005-06-23 at 12:05 -0400, Paul A Houle wrote:
> Rodd Clarkson wrote:
> 
> >I'm assuming you're talking about some sort of redundant raid (for
> >protecting data), and not striped raid (for speed improvements).  While
> >this is nice for situations where a HDD might fail, I don't see how this
> >offers security in terms of data damage based on viruses, worms, malware
> >or intrusions.  It doesn't matter how many 'redundant' drives you have
> >if they all carry (in essence) the same information.  Two buggers copies
> >of the same file are stiff two buggered copies.
> >  
> >
>     Yeah,  but the conventional choices for backup are unacceptable and 
> very few people will do them.

I'm not suggesting that this isn't a useful method to back up against
some sorts of potential data lose.  However, the discussion was about
data lost resulting for unwanted intrusion into you system including
hackers, worms, viruses and that sort of thing.  Given that all the
disks in a raid setup are part of a 'live' session, it's very unlikely
that damage to data will be limited to a single drive.  For example,
something that makes all your files zero bytes long will affect all the
disks in the RAID array.  So in this case a RAID array isn't going to
help protect your data.  RAID has it's advantages in protecting data,
but only when a single (redundant) drive fails.

My father in law used to back up his data on three disks (ZIPs - don't
ask ;-]) and one day he discovered that one of his important files was
corrupted.  The corruption had occurred some months earlier and what he
discovered was he had four copies of the same corrupted file.  Same goes
for a RAID setup.   Corrupt a file, and the file is corrupted on all the
disks, not just on one.  As such, RAID doesn't protect you against file
corruption, only against hardware failure.

regards


Rodd
-- 
"It's a fine line between denial and faith.
 It's much better on my side"