Lisa,

what about:
perform incremental backup from old box
generate backupset (devclass file, burn to CDs)
transfer the backupset to node
restore the data locally from backupset on new box
define *same* node name in dsm.opt
RENAME FILESPACE \\old_host\d$ \\new_host\d$
I have never performed such a trick but you can try it. Be careful with
Unicode-enabled filespaces if the client on old box was 4.1.x or below and
you put 4.2.x or 5.1 client on new box.

Zlatko Krastev
IT Consultant




Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
Sent by:        "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
To:     ADSM-L AT VM.MARIST DOT EDU
cc:

Subject:        frame relay nodes-- can I avoid another full backup?

Hello again,

here's the deal-

we have approximately 50 frame relay nodes that backup to TSM servers.
These NT boxes' drives are almost full (max is 16G on these) and
partitions
were created to hold data (D:\) and app shares (E:\).

We are getting replacements (80G and expandable), and will be xcopying
data
from the old to the new, but the OS will be 2000.  The plan right now is
to
just create a big D:\ drive and xcopy the files from the old D: and E:
drives to this new D: drive.  Host names may remain the same, but maybe
not-- since we won't be a full 2000 enterprise, and 2000 can't "emulate"
an
old NT BDC, the old boxes may remain in the remote locations to serve as
the BDCs-- without changing their hostnames.

How can I tell/trick TSM to not backup up these "new" files on a different
drive again and to know about previous versions of files with the old
drive
letter?  Or is this issue moot due to the OS change?  I can rename the
node
data and resolve the issue with a full backup due to the change in
hostname, but I don't see any way to rename the filespace names to keep
from having to do a full backup.  I am not up to speed on 5.1, maybe the
ability is there with the new code?


tia

lisa