If the drives are connected via SCSI directly to the filer, then a
reboot is required. If they are on FC, then sysconfig -t would rescan
the bus without a reboot.
Changing the drive order would require a change in netback config,
'cause now what used to be drive 3 on netbackup is now nrst1a on the
filer, so when a load command is sent to drive 3, the tape would get
loaded on drive 4.

We swap the drives physically in our library when we run into such
issues, so that the configs don't change on the netapp or netbackup.
But then we have FC connections from the netapp, so they don't need a
reboot.
Hope this helps
-G

On Tue, 14 Dec 2004 16:26:05 -0000, Khan Imran A
<Imran.a.Khan AT uk.fujitsu DOT com> wrote:
>  
>  
> 
> Hi 
> 
>   
> 
> This system has Windows 2000 backup server connected to two drives (1 & 2)
> in a L80 library. The other two drives (3 & 4) are connected to directly to
> a NetApps, robot is controlled by the windows server. One of the NDMP drives
> keeps going down to identify where the problem  we have swapped drives so
> drive 4 in the library connects to 1st SCSI channel on filer and drive 3
> connects to the 2nd channel previously it was the other way around. Is there
> a way of running something on the filer to refresh the SCSI bus without a
> reboot, so it sees the new drive order or does a reboot have to take place
> and secondly when the original configuration was done the drive 3 and 4 in
> NetBackup are configured as 
> 
>   
> (1)       filer1:nrst0a 
> (2)       filer1.nrst1a 
> 
>   
> 
> Since we have swapped the drives around would we have to change anything in
> NetBackup or is the path the same. 
> 
>   
> 
> And can someone remind me of the commands that can be run on the filer to
> retrieve the path of the drives as above 
> 
>   
> 
> Kind Regards, 
> 
>   
> 
> Imran Khan 
> 
>