Hi all,
we have a large number of users (university) and so a really large /home/
filesystem with a lot of folders. We want to give the users the chance to
backup their home-filesystem. So when they start the dsm-client
(aix-5.1.1-version) on the fileserver they have to wait a long time when
If you know which *EXACTLY* file you want to delete from TSM you can try
the following:
dsmc -virtualnode=RMAN node
dsmc expire the file in question
or
dsmc expire -filelist=a list
Later some tweaking of verd and reto may speed up the expiration.
Zlatko Krastev
IT Consultant
Please respond
I also had run into this a few clients back. I also had a problem open and
didn't get anywhere.
I've just done some 5.11 BMR restore tests and have not run into this - so
it may be fixed (or I just haven't run into the same situation yet.)
Tim Rushforth
City of Winnipeg
-Original
Guillaume,
there was a thread on this problem quite a while ago. Just search the list for delete
object 0.
Using TSM's SQL interface you can select the object-id from each file you want to
delete. Now you can create a macro file with line
like delete object 0 46364523. To delete the objects
Does anyone have some scripts they would like to share that perform backups
of DB2 on an NT/2K platform using TSM? I'm trying to wade through the DB2
manuals, but I need to get the backups going.
TIA,
Bill Boyer
DSS, Inc.
Hi Tab,
What are the impact o fusing a FILE rather than a DISK devclass
other than faster reclamation of copypool?
Thanks for the info.
Etienne
Jack,
It's been mentioned before, but if you use a disk pool for directories,
you should make the disk pool a FILE devclass.
When my
Next best choice is - stay where you are ! Or make some long-waited
upgrades for 20% of the budget - you will get more.
TSM stresses not the operating system itself (excluding TCP/IP stack) but
the box. Thus the question is formulated bad and you will get bad answer.
Ask yourself correct:
We
From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED]]On Behalf Of
Jin Bae Chi
We have TSM server installed on AIX 4.3.3 which is also the failover
server for 2 other production servers. All production clients are AIX.
We try to move TSM server to a dedicated one. We can get neither AIX nor
Just set-up a Storage Agent and have in mind that DB2 uses API Client not
B/A. So be careful in which dsm.sys dsm.opt you are changing options
(LANFREE ones for example).
Zlatko Krastev
IT Consultant
Please respond to ADSM: Dist Stor Manager [EMAIL PROTECTED]
Sent by:ADSM: Dist Stor
Gianluca,
if you have MSCS cluster set-up according to the TSM client docs you have
to define 3 nodes for active/passive and 4 nodes for active/active
cluster:
- Machine 1 - node A (local, small)
- Machine 2 - node B (local, small)
- Cluster group X - node X (shared, might be big)
- (optional)
Did you made the same change in the /usr/tivoli/tsm/client/api/bin? In
fact onbar through XBSA lib uses API not B/A client. Check DSMI_CONFIG,
etc.
Zlatko Krastev
IT Consultant
Please respond to ADSM: Dist Stor Manager [EMAIL PROTECTED]
Sent by:ADSM: Dist Stor Manager [EMAIL
You will not corrupt windows if restore properly - restore files, restore
systemobject. Files will get replaced with proper ones from systemobject
(unfortunately it gets bigger and bigger, ms putting too much garbage
there).
Zlatko Krastev
IT Consultant
Please respond to ADSM: Dist Stor
Hi Lindy,
unfortunately the attached data was not delivered. Could you just resend your message
with the script quoted inline?
Just some hints on Linux backup:
- check environment variables. Did you export the right settings?
- check the TSM servers log. Get the actlog console on your screen
Hi Guillaume,
We have been running the Oracle TDP for several years, and this has ALWAYS
been a problem. The backup objects are described by metadata BOTH in the
RMAN backup catalog and the TSM DB. Several things can cause the two DB's
to get out of sync, and then you have backup pieces left
From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED]]On Behalf Of
Jon Adams
Though I have successfully implemented the TSM Client solution on our
clustered servers, I do not appear to have the ability to restore them via
GUI. From the GUI, any cluster shared resources DO NOT appear
Instead of making the directory pool a FILE devclass, we have ours designed
in a hierarchy.
DIRCACHEDISK devclass and relatively small. Just enough for a nights'
backups. NextPool is DIRPOOL wich is a FILE devclass. During the daily
server storage pool backkups we:
BA STG DIRCACHE
In this case AutoVault *IS NOT* an alternative - usage of server-to-server
virtual volumes requires DRM license.
Zlatko Krastev
IT Consultant
Please respond to ADSM: Dist Stor Manager [EMAIL PROTECTED]
Sent by:ADSM: Dist Stor Manager [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
cc:
IBM 3590H was GA'd 07/02. The TSM device driver (adsmscsi.sys) can be used
on Windows. IBM device driver is required on other platforms.
Regards,
Hunny Kershaw
Internet Address: [EMAIL PROTECTED]
Tivoli Storage Manager - Device Driver Development
Tel (408) 256-2828, T/L 276-2828
Quality Policy
Hi TSM-ers,
any recommendations on how to backup DFS on a W2K cluster?
Should I use DFSBackupmntpnt YES or NO?
Is there a TSM manual that shows how to cope with DFS?
Greetings from Austria
Thomas Rupp
Vorarlberger Illwerke AG
MAIL: [EMAIL PROTECTED]
TEL:++43/5574/4991-251
FAX:
I've just done some bare metal restore tests with 5.11 client on w2k dc in a
test environment. 2 DC's in AD domain, native mode, integrated DNS, both
servers Global Catalog servers. All fsmo's on dc1, I restored dc2. The
restore worked fine except for TSM. Failed to start one of the services
Also, once you EXCLUDE these files, they are marked INACTIVE.
When you do a restore, by default you only get ACTIVE files back.
So even if you do a full restore and there are backups of some of these
files, you would have to really work at it to create yourself a problem!
-Original
Hello *SMers,
I have a couple of questions. We are possible moving to an LTO
drive in the near future. I am trying to determine how to do a TSM backup
without sending it to a separate cartridge. Our DBbackup is small and do
not see a reason to use such a large cartridge, do
Does anyone know if there are any know problems doing server-to-server with
the source server being a 4.2.x server and the target server being a 5.1.x
server? I'm trying to do a database backup and the 5.1 server keeps coming
up saying 'node (Windows) refused - node name not registered' even
23 matches
Mail list logo
