I understand your challenge, and I may raise additional questions for you to

Here is an option:
Configure and enable database backups (if not already done). You may
want to consider backing up the databases to disk on the disaster recovery

Backup the production server and all of its data to a specific pool of
tapes (this will help in isolating them for later use). You may want to
consider backing the master backup server to disk on the disaster recovery

Rename your disaster recovery server as the original, duplicate IP, DNS,

Use Netbackup official rebuild of the master server (this will bring
back all catalogs/images),

** you may want to restore any custom scripts at this time.

At this time you have a working master backup server that exist as a

Restore desired client (without the 2 phase import). Your time cost
here is greatly reduced

The key here is how current is your backup of the database. Worst case
scenario, that you must use an actual backup to retore the master server and
then after the reboot, retore the desired client.

If you have a server that will sit aside, but ready for disaster recovery, I
belive it would be easier to rebuild the server from scrach and the restore
the databases

Good luck, I hope this helps.

"Manpreet Singh" wrote in message
> We have setup a new environment for disaster recovery.
> The setup includes a new NetBackup master server connected to a
> media server which is connected to a StorageTek tape library. The tape

> has 6 tape drives.
> We just ran a test this week, were we tried to restore a
> database which had been backed up using the production backup
> environment. This backup spanned 11 tapes and took 9.5 hours to import

> restore. This is unacceptably long and we are looking for suggestions to
> reduce this interval to as small as possible.
> There may be more than one way of addressing this. For instance,
> increasing the number of drives in the library may be one
> of the option. However, we are more interested in a software
> solution. Is there a way for us to migrate the catalog from the
> production master server to the "disaster recovery" master server over the
> network and keep this updated on a fairly frequent basis going forward?
> By doing this, we
> could possibly skip the import steps all together. In the test restore we
> ran, the import phase took 5 hours.
> The Netbackup support said that copying catalog files is not supported.
> No surprise here.
> Thanks.