Hi Timothy,
this was already discussed in this list...
the RC=4 in 4.2.0.0 client is a code defect.
You should use the last fixtest for 4.2.
--
regards / Mit freundlichen Gruessen
Joachim Stumpf
Datev eG
Nuremberg - Germany
taford [EMAIL PROTECTED] (Dienstag, 25. Juni 2002, 18:49:30,
Are you going to use RMAN with Oracle (Required the TDP for Oracle) or are
you going to do Offline Backups?
Paul D. Seay, Jr.
Technical Specialist
Naptheon, INC
757-688-8180
-Original Message-
From: Rupp Thomas (Illwerke) [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, June 25, 2002 2:40 PM
I will not explain to you what Veritas bpVault is then. It might put you 6
feet under. Looks like Legato and Veritas are racing for the award, Joke of
Backup Solutions.
Paul D. Seay, Jr.
Technical Specialist
Naptheon, INC
757-688-8180
-Original Message-
From: Peter Pijpelink -
Hi all,
a question about increasing the performance of my TSM-Server:
TSM-Server : IBM H80 ( 2-way RS64III-450MHz-processor, 2GB memory ), AIX
4.3.3 , TSM 4.1.4
I see with the monitor-tool that my server has over a longer period a
cpu-utilization of more than 80%. This server only handles TSM,
Hi All
Have you seen this one before.
In a Windows 2000 SAN based environment, the BA Client GUI only shows local disks that
can be backed up but not SAN based disks.
Doing a selective backup from the command line works fine which indicates that the
Storage Agent, drive mappings etc, are
Just another quickie - does anyone have any experience of image backup
of a Windows 2000 machine? If so, how does it work when doing a
restore? How can you do a restore if your W2K machine crashes
completely? I have instructions for creating an image backup but
haven't seen any useful info on
Hi Folks,
I have a question regarding DBBACKUPTRIGGER.
First I will set the scene. We have ADMIN SCHEDULES defined to execute a
FULL DB BACKUP weekly, and on a DAILY basis we take two INCREMENTAL DB
BACKUPS, again using ADMIN SCHEDULES.
Our DBBACKUPTRIGGER is defined as follows :
Actually you can reconstruct the aggregates during a move data. I forget
what version it started with but there is a command line option for move
data called Reconstruct=yes that will reconstruct the aggregates during a
move data
Mark
-Original Message-
From: Zlatko Krastev/ACIT
IMHO,
you already have a decent box, I wouldn't think that CPU could be an issue. Is
user+system 80%? or is iowait 80%? This is an important difference.
Depending on file caching, more memory is always a good upgrade.
Then I would look at your I/O adapters, SCSI, SSA, etc. Adding more is
I was reading the TSM Admin guide where it recommends that you periodically
do a reorg of the database. We have had our TSM server running since 1998
and have never done a reorg. Our database is around 65G. We're running TSM
4.2.1.11 on OS/390.
Thomas S. Ritter
Promedica Health System
We had a client that had high CPU usage due to some TSM database
corruptions. The CPU utilization was constantly high, even when the server
had no sessions or processes running. Now, this was an NT box, but once we
ran some AUDITDB's with FIX=YES their CPU went back to normal. They were
also
Use sar -P ALL ,look at each cpu how they are utilized!!
Balanand Pinni
SBC Services Inc.
Work:314-206-5911
Pager:1-800-451-6897
Email ID :[EMAIL PROTECTED]
[EMAIL PROTECTED] e.mail pager
-Original Message-
From: Christoph Pilgram
[mailto:[EMAIL PROTECTED]]
Sent: Wednesday, June 26,
There's a third way - you can use reclaim. This is similar to move data,
with reconstruct=yes, but is more automated. Just set reclaimstg to your
new storage pool, and reclaim down to 1. You may have to do a few move
datas at the end, to get the tapes that are 0% reclaimable.
At 08:05 AM
You didn't indicated if this filesystem will be backed up weekly or daily
with those rules. My answer assumes weekly, or daily with a small
filesystem (not too many files, and too many isn't easily definable).
Don't backup that filesystem (just remove it from domain statement, or
exclude.fs it),
Your CPU utilization profile does not match mine so, I do not know if I can
help.
We have a P660-6H1 identically configured, except I use Shark Disk and
Magstar tape using FC and the Ethernet is Gigabit. The only time I see high
CPU utlization is during backups coming down the ethernet. At
Just a little comment on #2...the pre/postschedule commands run every time
you run a schedule against that node! So if you were to maybe run a
CLIENTACTION against the node to do a special backup of a file/directory,
not necessarily the whole server, you're shutting down your database. If
that's
Don,
if you allow me I prefer Jack's order - Sun or even HP-UX before WinNT/2k
if any mentioning of SAN is made.
The way Win2k deals with SAN zoning, naming or any
addition/removal/reconfiguration in the fabric is just terrible. I had no
chance to experiment with XP or (God help me) .Net but to
I have done it in test, but not production til next week. Out DB went
from 46GB to 26GB. I was able to do selects from the database that
would never finish on the prod system. I obviously have no idea of the
long term effects are.
Andy Carlson|\
I have one more question about the unloaddb/loaddb procedure. When
you do the loadformat command do you count the
mirrored logs in the count. For example if I have 2 recovery log files
and copies of the 2 in another directory I would assume
that my loadformat command would have 4
As a TSM newbie I'd be grateful for some hints and tips - I'm not after
instructions because I have those!
A customer has a small office, 15 or 20 people with laptops (20 Gb HDDs
mostly) and a new Windows 2000/TSM server that contains 4x 60Gb IDE
disks. They have no tape robot.
What is the
The answer to your question is yes.
I forgot my machine has 4 processors in it.
Your CPU utilization profile does not match mine so, I do not know if I can
help.
We have a P660-6H1 identically configured, except I use Shark Disk and
Magstar tape using FC and the Ethernet is Gigabit. The only
I have an NT 4.0 client running TSM version 3.7.2.19. Server version
AIX4.3.3 running 3.7.5.0. The problem is the backup stalls and eventually
dies on one of the drives. All other drives complete. I have changed the
idletimeout on the server, no luck. No error messages are generated. A
chkdsk has
I am suddenly having the same problem...
Reclamation is very slow. I barely manage to reclaim 1 GB/hour, which
wasn't the case in the beginning (I think after all in the beginning I
didn't have so much data outside). I update the stg to reclaim=60. I
have a diskpool on a SAN (FastT500) and
Hello everyone,
I have a TSM/AIX server at v4224 level. Server has 5 GB of memory. TSM server's
log disk size is 10 GB. Two days back I changed the bufpoolsize from 1 GB to 1.5
GB, and logpool size from 2048 to 4096 and recycled the TSM server. Since then I
have lost sleep. TSM server is
Could someone explain what the 'aggregates' are that the
'reconstruct' parameter addresses in the move data command for me?
... Thanks ... Jack
-Original Message-
From: Paul Zarnowski [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, June 26, 2002 9:05 AM
To: [EMAIL PROTECTED]
Subject: Re:
Hi
Can anybody explain to me those output..
1. Running a Q OCC system_1 got:
Node NameType Filespace StorageNumber of
Physical Logical
Name Pool Name Files
Space Space
OccupiedOccupied
We had that happen to us when we used the 4.2.1 client.
Now we are using the 4.12 client and it is better.
Marc Levitan
Storage Manager
PFPC Global Fund Services
Mubashir
Farooqi To: [EMAIL PROTECTED]
The mirrors are not counted. You have to remirror it once it is loaded.
Andy Carlson|\ _,,,---,,_
Senior Technical Specialist ZZZzz /,`.-'`'-. ;-;;,_
BJC Health Care|,4- ) )-,_. ,\ ( `'-'
St. Louis,
Hi,
Our data servers(winnt/win2000) are overloaded. We want to archive data, to
free up some space, based on last access time. Is there a way to automate
this, using the native ba client. Can it be done use a script? We know of
the OTG software, but at the moment that can't be used.
Thank You
Here is the newly minted APAR number for the problem I experienced
last week:
IC33977
This was the problem that prevented me from upgrading to v5.1.x and
requires a database audit.
It was too late to get everyone's agreement last Friday to run
the audit, so it will happen over the July 4th
And to continue...
If you moved the data without reconstructing the aggregates the process
would move these 'bundles of files' whole, with the empty spaces that
occurred due to expiration, and the resulting output tape would have
reclaimable space and not be 100% utilized.
Mark
-Original
You ever been an MVS person ? Used partitioned data sets ? ? ?
OK, TSM will do a trick where it bundles a lot of little files in a big file
within ~TSM space~.
OH, look at it like a ZIP file
so if a bunch of little files are in this one big file that TSM is dealing
with, natrually some files
If running from a schedule, check the schedule log to see if always stalls on the same
file. If so, exclude that file and try again.
-Original Message-
From: Blair, Georgia [SMTP:[EMAIL PROTECTED]]
Sent: Wednesday, June 26, 2002 7:48 AM
To: [EMAIL PROTECTED]
Subject:
Date: June 26, 2002 Time: 12:34 PM
From: Jerry Lawson
The Hartford Insurance Group
860 547-2960[EMAIL PROTECTED]
-
I have seen some strange behavior with
Andy, thanks for the response. How long did the whole process take?
Thomas S. Ritter
Promedica Health System
Technical Coordinator
(419) 291-0049
-Original Message-
From: Andrew Carlson [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, June 26, 2002 10:29 AM
To: [EMAIL PROTECTED]
SAN (aka LANfree) protocols are good for moving a few huge files. If you
have lots of little files (as is common in W2K) then you are better off with
TCP (what your LAN probably uses).
Have fun!
Kai.
-Original Message-
From: Mike Brown [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, 26
Why my LAN free backup even slower then my LAN backup?
I got really bad performance, 1.2MBps for LAN, 900KBps for LANfree Backup.
My TSM server is Version 4.2.1.11 in AIX 4.3.3, p640, 1 power3 processor, 1G memory;
My TSM Storage Agent veriosn 4.2.1.11 on Win2k Advanced server, TSM client
When the primary location of a file is in a primary DISK random access
storage pool, then this is working as designed. (That's the answer we got
from Tivoli) In this case the reclamation task processes 1 file at a time
per transaction. When the primary storage pool is a SEQUENTIAL media (either
This was from an old email of mine. We got info that my last paragraph
was a good guess, it is a TSM design problem. Notice carefully if when
the Reclamation is running slow, that there is just an output tape
and no input tape. This means it is getting input files from disk and not
tape, if
More than likely, you have a situation where the directories are being bound
to the longest mgmt class you have for archiving in that policy domain.
If you look at the regular backup data the same thing should be occurring.
Directory entries are showing but the data may already be expired. The
Hi.
I don't see much return on making those primary disk storagepools
sequential, because once they get tape hardware, you can just move data or
migrate the backup data off of the random diskpool volumes. In fact, it's
more of a headache because you'll have to start reclaiming them and whatnot.
Mike
It depends of what your configuration looks like and what files you are
backing up.
Think of this:
If you are backing up many small files on one SAN based machine and big
files on another machine, it goes without saying that the machine backing up
the bigger files will have better
also check on the cleanup archdir command...
it will get rid of all those extra entries
CLEAN ARCHDIR node_name {DELETEDUPLICATES | SHOWSTATS | RESETDESCR |
1DELETDIRS } [FORMAT=S|D] [WAIT=NO|YES]
Dwight
-Original Message-
From: Lawson, Jerry W (ETSD, IT) [mailto:[EMAIL PROTECTED]]
Has it left the treasury yet?
Paul D. Seay, Jr.
Technical Specialist
Naptheon, INC
757-688-8180
-Original Message-
From: Gretchen L. Thiele [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, June 26, 2002 12:19 PM
To: [EMAIL PROTECTED]
Subject: Re: TSM v4.2.2.5 to v5.1.0.0 upgrade failed
Actually, I use the same Win2k host, all backup direct to tape, all no disk pool
involed, backup same files (backup whole C:\ drive, delete the filespace before it run
next time backup).
To be more precisely, my question should be Was the 900KBps is the expected
performance in this LAN free
I just finished a TSM server upgrade from 4.1.3.2 to 4.2.0.0. The server
is on AIX 4.3.3. I did some testing afterward and found out that the data
from my AIX TSM 4.1.0.0 Clients doesn't appear to be there anymore
(Archives, files, etc.), in other words, I can't see or access it (it was
H.. try using namedpipes instead of tcpip for your commmethod, and
see if that makes any difference.
Regards,
Andy
Andy Raibeck
IBM Software Group
Tivoli Storage Manager Client Development
Internal Notes e-mail: Andrew Raibeck/Tucson/IBM@IBMUS
Internet e-mail: [EMAIL PROTECTED] (change
On out test box, it took 16 hours to unload, and 7 hours to load. I
started an unneeded audit, just to see how it would go, and I estimated
it would take 92 hours.
Andy Carlson|\ _,,,---,,_
Senior Technical Specialist ZZZzz /,`.-'`'-.
Hi there
Here is our setup. We use Gresham's DistribuTAPE software to manage our STK 9840
drives, which are in an STK 9310 (BIG mainframe library).The drives are fiber attached
through a Brocade silkworm switch.
We originally used Gresham's AdvanTAPE driver with the generictape device class.
Hi *,
pls disregard...
issues resolved, thx.
Lindsey Thomson
Hi,
Sorry for the long post but I see posts with too little data and
subsequent responses requiring more data AND the correct data to debug
this issue may not even be in this post!!!
Pls keep in mind that I have 27 AIX client's running, successfully, and
ALL client's have been configured
We are currently running into some expiration issues on our current server
of 4.2.1.7. It is not deleting files due to some archive files we have.
Therefore, Support said to run an expire with expire inv q=n w=n
skipdirs=yes du=60 every hour till we get the database down to a managable
I did try it and it does move a lot faster.. So this has nothing to do
with the fact that my diskpool is on a SAN drives right? So keeping data
on diskpool is not a good idea in small installations? I was told by
Tivoli techs to try and keep a couple of days of backups on disk if
possible to
I just ran into something similar on TSM server 4.2.1.13 and haven't opened
a PMR on it yet -- my admin password expired, and I got the authentication
failure message. No word about 'password expired', and NO option to enter a
new one. Maybe your client password has expired? Try q node nodename
The server still hasn't come back online - 125 minutes of log churning.
Any ideas?
ANR2017I Administrator issued command: IMPORT NODE DLECU filespace=/var
filedata=archive devclass=FILE_CLASS
volumename=FILE:/adsm-tmp/ADSM.S5123069.EXP
ANR2345E IMPORT NODE:
Yes Bill,
Your point is correct if the pre/post commands are coded in dsm.sys. However is
possible to specify them on the command line and hence through the -opt parameter of a
schedule, in which case they only apply to that schedule.
Steve.
[EMAIL PROTECTED] 26/06/2002 22:53:03
Just a
I'm getting something similar.
The environment is Veritas cluster in which nodes keep moving from one machine to
another. Everything worked fine at client level 3.7, but at 4.2.1.0 and 4.2.1.28
(server is 4.2.1.8) I keep getting these errors. My nodes are password generate with
a long
It has more to do with LTO tapes becuase they don't handle the
little files spoon fed one at a time as well as say 3570 tapes.
I have both here and can compare. The root cause is TSM design
though. (I still like LTO, it is really great in my view! Now that I
understand what's happening, I just
Folks,
I have TSM server 4.1.4 on a Sun E450 with 4 processors, 2Gb of ram, 21
disks as the primary stgpool across 3 SCSI busses. We have around 100
nodes backing up to this puppy, backing up around 90Gb nightly (
achiving 11Gb a night ). I'm seeing pretty consistent throughput from
the nodes
Gretchen,
Have you determined whether your problem is unique to Princeton, or something
that could effect all of us. I was going to go from 4.1.5 to 5.1.1 on
Wednesday until you reported your problem. I'm retesting the upgrade procedure
now, but I'd really like to know if I should plan on
60 matches
Mail list logo
