Restore files without directory listing

csops

ADSM.ORG Member
Joined
May 5, 2009
Messages
49
Reaction score
0
Points
0
Hi,

I have a bit of an issue with an urgent restore. It seems that the directories belonging to a node have got bound to a MC that is different than the files. It seems that the directories have expired before the files, however I would the expect the files to be still retained based on the copy group policies.

Can I actually restore there files if I do not have directory structure/listing? If so, how do go about this? I also don't know the exact names of the files!!

Regards,

CSops
 
Hi,

in these cases you have to use the CMDline interface (as GUI client cannot display those files - there was a IBM technote about this - will look for it) - use standard commands "q backup /rootpath/* -sub=yes" and "restore ..."
IF it (for any reason) does not work, you can get the exact file names using "select ...." commands from the server side - but can be time consuming.

Hope it helps

Harry
 
Hi Harry,

Thank you for your reply. I have had no luck digging out that IBM Tech Note. It would be great if you had a link. I'm finding it very hard to locate the files. They should still be there given my policies.

I've played around with the CLI using q backup statements but I cannot see the files (and directories) that I'm looking for. I've using commands such as the below to look for all files with Database in the name.

q b -ina -sub=yes "\mypath\*Database*"

How would I use the Select statements on the server side. What type of statements would I use?
 
Hi,

from the server side (your nodename is "NODEX"):
q files NODEX
Code:
tsm: TSMSRV_TTC>q files nodex

Node Name           Filespace       FSID     Platform     Filespace     Is Files-        Capacity       Pct
                    Name                                  Type            pace               (MB)      Util
                                                                        Unicode?
---------------     -----------     ----     --------     ---------     ---------     -----------     -----
nodex             \\nodex\c$         4     WinNT        NTFS             Yes           81,909.5      54.5

so you have the filespace name/ID

Then you can run something like (use filespace_name OR filespace_id):
Code:
select node_name,hl_name,ll_name from backups where node_name='NODEX' and filespace_name='\\nodex\c$'

You can refine the output by using additional parameters (hl_name is partial path starting from the filespace level), using LIKE statement etc.
THIS CAN TAKE LONG TIME

If you cannot find it this way, TSM does not have it ....

Harry
 
Thanks Harry. This is the statement that sprung to my mind. There are over 2 million active files in this filespace! It will take a long time!!!
 
Back
Top