The simplest way would be to manipulate your Server Parallelism and Target
Sessions per drive such that the last drive would not be needed.  For
example, if your Target Sessions are 4, set Server Parallelism to 12.  Or
increase Target Sessions to 8 and set Server Parallelism to 24.  This way,
the server should have no need to use the last drive.  If you have Storage
Nodes, this might not work so straight-forwardly.

I always try to avoid restricting pools to drives - increases management
complexity.  Also, another poster suggested that the unused drive would
still be available for restores, but I don't believe that would work - in
my experience, if a drive is not eligible for a pool, it won't mount a tape
for that pool even for a restore.

On Fri, 14 Mar 2003 13:14:04 -0800, Matthew Braithwaite <mab AT CNET DOT COM>
wrote:

>Is there a way to do one of the following?
>
>  (a) Tell Networker to surrender a drive at the next available
>  opportunity?  By surrender, I mean something like ``do not use the
>  drive for any operation except one that I start by hand''.
>
>  (b) Tell Networker not to use a drive at all, except for certain
>  purposes -- mainly inventories and restores.
>
>The problem I'm trying to solve is that all N drives in our jukebox
>are busy constantly.  We could still get all our backups done with N-1
>drives, though: the server is network-bandwidth constrained, rather
>than tape-bandwidth constrained.  So we'd like to have a drive sitting
>idle for the occasional restore and for jukebox inventories (which
>require a free drive, even if that drive is never needed to read a
>label).
>
>We're using Networker 5.5.1 on Solaris 7, and a Sun L1000 four-drive,
>30-slot DLT jukebox.
>

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=