ZFS (was Reliable Firewire drives)

[Alex Satrapa]

On 13/02/2008, at 02:26 , Dan Shoop wrote:
> Neither one should have occurred had only the sysadmins been  
> exercising any modicum of due diligence.

Even if only one had occurred, the problem would not have presented  
itself.  Doing a file-by-file copy means you insure yourself from  
errors in the source file system.  Doing a verification restore means  
you can be alerted to the fact that your backup copy doesn't work,  
which will give you the opportunity to fix the underlying problem  
before you have to rely on it.

But I agree with the principle of your statement.

To everyone else on this list, take the original story as a warning -  
if you're doing backups, you really do need to restore from the  
backup to a scratch drive, and verify that the recovered copy is  
close enough to the original to be of value to you.

It's worth remounting filesystems with "access time" turned off for  
the verification part, to save about up to 50% of your time ;)

> Your tools can only do their tricks if you keep them in working  
> order. I don't expect a broken hammer to drive nails and shouldn't  
> complain when its head flies off as I try to drive a nail and hits  
> me in the face.

Especially not when the vendor has already warned you that the head  
will fly off.

Alex