We take the tapes out of the library and set the access to Offsite and the
location to Vault. Access of Offsite prevents TSM from trying to mount
the tape, and location is a free form text field, so we just call it Vault
as the tapes our stored onsite in a Vault. Once a week we take tapes
Yahya,
If it is within the same domain, you might have trouble.
You also didn't state whether you were running
Exchange 5.5 or Exchange 2000. They each have their
own issues involved with alternate server restores.
Keep in mind there is the /FROMEXCSERVER=servername
option to allow TDP for
Hi All
I have a compaq Proliant 370 with 2 750Mhz CPU's With 2Gig of Memory and
2 1Gigabit Network Cards, teamed with 1 active and the other is a
failover standby card. Does anyone have some info on how best to setup
these cards for the best performance. I have a number of errors where I
get
BTW: you are having an exposure with this bring-back + back-offsite. While
the box is onsite you have no off-site copies and are vulnerable to
disaster.
TSM is made to handle off-site copies without any need to bring them
on-site before new copy is send out. Look again through your procedures.
Well...
Using realistice daily change rates, real hardware, high performace
drives and some tricks, no problem
On dinsdag, november 19, 2002, at 06:06 , Dan Foster wrote:
2 PB is 2,048 TB, or 2,097,152 GB.
--
Met vriendelijke groeten,
Remco Post
SARA - Stichting Academisch
-- Q1 - We are considering whether to set TSM Client Compression=Yes ...
I've installed client with node compression for a customer with 600+ GB
Oracle DB backed up off-line. The node was p660-6M1 and result was 10-11
MB/s uncompressed 3-4 MB/s compressed per thread/per processor
I ship between 8 and 12 LTO tapes off-site every weekday morning. Once a
week we generate a pull list of all off-site tapes in pending or empty
status and have them (plust the oldest week's TSM database backups) brought
back. This runs to about 40 tapes coming back.
I do daily reclaims of some of
ADIC has some pretty high-end libraries. Their AML/2 library claims a top
capacity of over 5PB, 400 drives, and over 7,600 pieces of media. It
supports around 20 different media types. When SAIT comes out, that
capacity would likely go up significantly higher.
At 05:51 PM 11/19/2002 +, Dan
Good Morning TSM'ers,
Can someone tell me where the activity log maps to on the AIX side.
I would like to expand the duration of the log, but I need to know how much
is being utilized currently before I do that.
Thanks in advance
-kane
The actlog is another table in the TSM DB. Do a q status and
Activity Log Retention Period will show how many days you are
keeping. I would keep a minimum of 7 days, I keep 30 days. It can
be very useful in tracking down problems.
To change use set actlogretention.
David B. Longo
System
Well, you expand the duration of the activity log by issuing command Set
ACT 30 This wil keep activoty log for 30 days. To see how long its
retained for now, type command Q ST' and look for Activity Log Retention
Period:
Thanks,
Robert Rippy.
From: Nelson Kane [EMAIL PROTECTED] on 11/20/2002
If you're concerned about your activity log eating up too much space in your
TSM database, don't be. It uses relatively tiny amounts compared to the
contents table. But there'e probably no good reason to keep more than 30
days.
-
Mr. Lindsay Morris
Lead
I think the activity log is normally held inside the TSM database not as a
separate file hence increasing the length of time you keep it for will
impact on your database volume usage.
Regards
Tim
-Original Message-
From: Nelson Kane [mailto:[EMAIL PROTECTED]]
Sent: 20 November 2002
Greetings All,
Any idea what could cause TSM to error out ?
f1n15 /tsg/tsmmore dsmerror.log
19-11-2002 06:00:01 cuSignOnResp: Server rejected session; result code: 52
19-11-2002 06:00:01 sessOpen: Error 52 receiving SignOnResp verb from
server
20-11-2002 06:05:15 ANS1228E Sending of object
Have you checked if your password is still valid?
If I am remembering right then 52 notes that your password has expired.
regards, Manuel
- Original Message -
From: shekhar Dhotre [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Sent: Wednesday, November 20, 2002 4:12 PM
Subject: Server rejected
I have collocation switched on our onsite tapepool at the moment, what are
the effects of switching this setting on and off.
I'm guessing that when switched off the volumes that are 'filling' will be
eventually used to 100%?
What will be the effects of then switching collocation back on? Will
h ..
If I start dsmc incr session , TSM won't ask me for password ,it
starts backing up .
Normal File-- 1,431 /etc/vfs [Sent]
Normal File-- 101 /etc/xtab [Sent]
Normal File-- 1,310 /etc/xtiso.conf [Sent]
Directory-- 512
Thank you all for the quick response!
If collocation is on now, then each tape belongs to a specific server. If
you turn it off, then TSM gets the least filling volume first to put data
on even if the data belongs to another server. Makes for more efficient use
of the tapes but then you could have data from multiple servers on one
We keep ActLogRetention, EventRetention and SummaryRetention at 37 days.
Reason: we generate some monthly reports from those tables. That gives
us a 6-7 day window to generate the reports, plus extra time to regenerate
them if something is amiss, and to allow for long weekends, vacations,
thanks!
Does anyone have a good sound and understandable Backup Request Form that
they use for their users or clients. I need revise the one we are using
because it is to complicated for the average user.
Thanks,
Robert Rippy
I would look at the number of running scheduled sessions at that time. The
message is pointing at a server action (server rejected session), so try the
actlog.
-Oorspronkelijk bericht-
Van: shekhar Dhotre [mailto:[EMAIL PROTECTED]]
Verzonden: woensdag 20 november 2002 16:28
Aan: [EMAIL
Error 52 isn't a password prob. It looks like TSM server didn't respond
fast enough to the client and the session was rejected. Try increasing your
timeout limit. Could be a lot of stuff hitting the server at that
time...aka backups. and the TSM Server is overloaded.
Thanks,
Robert Rippy.
Phil, please clarify.did you completely replace the B1A drives
with new E1A drives, or did IBM convert your existing B1A's to E1A's with
an upgrade kit?
James, Phil [EMAIL PROTECTED]
Sent by: ADSM: Dist Stor Manager [EMAIL PROTECTED]
19-Nov-2002 18:10
Please respond to ADSM: Dist
When I try to create a scheduler service on a NT4 running TSM Version 4.2.31
I get this error:
The procedure Entry Point TSMEnumDependentServiceEnd Could not be located in
the Dynamic Link Library TMSUTIL1.DLL
Please help
Thanks
Mehdi Amini
LAN/WAN Engineer
ValueOptions
12369 Sunrise Valley
Check ur network on both sides and do ftp to check xfer rate .
Else change ttl value for tcp/ip.
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: Robert L. Rippy [mailto:[EMAIL
Manuel,
you are right , actlog shows that password has expired . in dsm.sys file I
have passwordaccess=generate .
Where is the password parameter? so that I can reset it ? when I add new
node in tsm ,I do dsmc on client and enter password for the first time and
tsm stores/uses it .
Greetings,
I have a customer with a unique challenge. Maybe it's not, but it
seems to be to me because I am not a Novell guy. They are considering
implementing TSM in their environment to back up a bunch of Novell
servers that are out in remote plants with real slow links, like 256KBs.
The
We have the FileExit set as a receiver and trim it everyday to datestamped
files. The filesize averages between 1 and 2 MB per day.
We've got 60 nodes, 5 Storage Agents, 16 drives and too many storage pools..
(another story)
You will need to alter the dsmserv.opt file to enable this FileExit
That doesn't look good. I would recommend calling support at this point.
--
Joshua S. Bassi
IBM Certified - AIX 4/5L, SAN, Shark
Tivoli Certified Consultant - ADSM/TSM
eServer Systems Expert -pSeries HACMP
AIX, HACMP, Storage, TSM Consultant
Cell (831) 595-3962
[EMAIL PROTECTED]
you might try reloading your client code.
Ted W. Qualls
UNIX Enterprise Engineering
PepsiCo Business Solutions Group
5080 Spectrum Drive
Suite 600W
Addison, TX 75001
office: 972.376.7809
pgr:800.946.4645 pin 1090913
cell: 469.682.1773
email: [EMAIL PROTECTED]
-Original
Have in mind after node compression your 10 GB
files will shrink to approx. 4GB and diskpool might be sufficient.
We tend to get 90 % compression for a lot of our Oracle data. So 10 GB may
shrink to 1 GB. Give it a try to see what compression rate you get.
I would uninstall the client, reboot, then search for any/all occurances
of the TSMUTIL1.DLL. If you find any...DELETE 'em!
Then re-install the client.
Bill Boyer
DSS, Inc.
On Wed, 2002-11-20 at 14:38, Qualls, Ted W {PBSG} wrote:
you might try reloading your client code.
Ted W. Qualls
We had this happen to us also. A Win-something server worked fine until
the TSM client software was upgraded. The E: drive stopped getting backed
up. We noticed and IE error when starting things up so the box's admin
upgraded IE and the problem went away.
I can get more details from our Win/NT
we run a 3494 tape library with 3590 tape drives. after our backup are
finished, we run a backup stgpool to a copy pool on about 1.6TB of data. as
you can imagine, even with 8 drives this takes some time. are there any
server, device or other parameters we can tune to improve the tape-to-tape
I'm a UNIX guy who just got strapped with backing up Exchange. Installed
TDP for Exchange, tried a backup, got this in my actlog. Haven't found a
reference to this yet, anyone know what this means? The client says
Requested data is offline, but everyone is accessing their mail.
Thanks,
Hi All,
A few questions pertaining to nfs mounts and using both scheduled
backups as well as dsmc incrementals. When using ./dsmc -incr
/var/backup/20021120/FullBackup_a.log where the file is on an nfs
mounted partition do I need to worry about adding the domain /all-nfs
parameter? I would
My first guess would be re-install the client code. If you downloaded it
from the web, download it again.
Good luck!
John G. Talafous IS Technical Principal
The Timken CompanyGlobal Software Support
P.O. Box 6927 Data Management
1835 Dueber Ave. S.W.
Also might want to check your ~PATH~ and if TSMUTIL1.DLL exists somewhere in
that path.
Might be it can't find the entry point because it can't find the file to
begin with :-O
Dwight
-Original Message-
From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED]] On Behalf Of
Amini, Mehdi
The storage pool backup that you are doing could be directly affected by the
database performance. Is this 1.6TB a lot of small files?
On the IO configuration side, how are your tape drives connected, SCSI or
FC? I suspect SCSI. If you have more than 2 drives, on a SCSI adapter that
is the
my knowledge in Cisco devices also is very far from expert level I so
consulted with a colleague. His answer was the same - there is no Cisco
device to mess with your TCP stream. You can have a switch (at OSI layer
2) which deals with Ethernet frames, a router forwarding IP packets or L3
AFAIK volumes are used in round-robin. Only if they are of different sizes
small ones can filled and not be used lowering the parallelism.
Zlatko Krastev
IT Consultant
Andrew Carlson [EMAIL PROTECTED]
Sent by: ADSM: Dist Stor Manager [EMAIL PROTECTED]
07.11.2002 19:33
Please respond to
Sias,
You were right. Thanks!
The scheduler service was logging in as Local System account. I thought
that account automatically had permissions to the file system, but
apparently not. Yesterday I gave the local System group full control of
the drive that wasn't backing up and last night the
You can find long discussions on this topic in the list archives.
- it is mostly disadvantageous to have more than one or two DB volumes per
disk/array - parallelism you create with more volumes results disk heads
moving back and forth. You are shooting yourself in the leg.
- RAID 5 is definitely
Neil,
This sounds like your backups are being directed to tape.
You need to indictate that you are willing to wait
for tape mounts. Try this command:
TDPEXCC SET MOUNTWait=Yes
Then issue
TDPEXCC QUERY TDP
to make sure the setting is set to YES.
At that point, retry the backup.
If it
I have a 40gb database on RAID5 with no performance issues.. fyi...
not sure if you consider this small, average, or heavy.
-Original Message-
From: Zlatko Krastev/ACIT [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, November 20, 2002 3:02 PM
To: [EMAIL PROTECTED]
Subject: Re: Database
Take a look at the movebatchsize, movesizethresh and bufpoolsize. These
3 options have a direct corrolation for tape to tape copy processes.
--
Joshua S. Bassi
IBM Certified - AIX 4/5L, SAN, Shark
Tivoli Certified Consultant - ADSM/TSM
eServer Systems Expert -pSeries HACMP
AIX, HACMP, Storage,
Is 1.6 TB the amount of must have/critical information ?
and is that already compressed ?
Knowing each tape mount runs about 90 seconds, any way to reduce tape mounts
will speed up copies.
If you have collocation on, turning it off ~could~ help... and collocation
can/is set on both primary pools
Actually, there are 2 speeds to be concerned with.
Data transfer from the host to the 3590 controller.
SCSI, depends on length and type of connection (34 was the top on
SCSI E1A).
40 MB/sec on FC E1A advertised, but I have seen higher (FC)
70 MB/sec on FC H1A advertised
thanks for that insight. i forgot to mention we are running fibre channel
and not scsi drives. yes, the data is must have... it is all production data
and is being copied for DRM.
how much of a performance hit will it take to turn compression on at the
client level?
-Original Message-
Hi, TSMers,
I'm managing 2 TSM servers on AIX now. When I installed TSM client on
my desktop WindowsXP, I could install also command line admin. But, I
can only use it for one server. Is there anyway that I can specify 2
different servername or TCPIPaddress for servers so that I can start
Hi all,
it's possible Tivoli web admin works on https?
Thanks All
Estanislao Sanmartín Rejo
[EMAIL PROTECTED]
Does the H1A use a different tape then the E1A?
---
Bruce Kamp
Midrange Systems Analyst II
Memorial Healthcare System
E: [EMAIL PROTECTED]
P: (954) 987-2020 x4597
F: (954) 985-1404
---
-Original Message-
From: Seay,
You can launch dsmadmc with the -tcpserveraddress and -tcpport settings to
point to a different server, i.e.:
dsmadmc -tcps=server2
dsmadmc -tcps=server2 -tcpp=1510
etc.
Regards,
Andy
Andy Raibeck
IBM Software Group
Tivoli Storage Manager Client Development
Internal Notes e-mail: Andrew
Add another server stanza in dsm.opt:
servername other
tcpserveraddress other.dns.name.or.ip
then you can say
dsmadmc -se=other ...
-
Mr. Lindsay Morris
Lead Architect, Servergraph
www.servergraph.com http://www.servergraph.com
859-253-8000
Yes, there is almost certainly a mismatch between the executable file (not
sure if this is dsmcutil.exe or dsm.exe) and the tsmutil1.dll file. Check
the modification dates for these files, and see if the timestamps match.
I'll guess that tsmutil1.dll is older. Note: tsmutil1.dll is located in
the
I run a cron job every midnight that dumps the previous day's actlog,
status, and a few other interesting figures to a text file. I keep two
year's worth. As David said, they are useful in tracking down problems.
- Kai.
A life spent making mistakes is not only more honorable, but more useful
Yep, even better than my suggestion!
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 eye to i to reply)
The only dumb question is the one that goes unasked.
Is anyone using some sort of script to verify their backups by a byte total?
I've got scripts to monitor events and the dsmerror and act logs but still
miss files.
Some time ago we designed a Bourne shell script to do backup/restore
verification for Unix systems, using a verification machine with
No, they both can use either the J (standard) or K (extended length)
type tapes.
--
Joshua S. Bassi
IBM Certified - AIX 4/5L, SAN, Shark
Tivoli Certified Consultant - ADSM/TSM
eServer Systems Expert -pSeries HACMP
AIX, HACMP, Storage, TSM Consultant
Cell (831) 595-3962
[EMAIL PROTECTED]
Is it possible to run two TSM clients on Solaris (2.8) with them
each having a scheduler daemon running? If so, I'd appreciate any
information on it!
I'm not sure whether 'two TSM clients' means two sets of client code
at different levels, or two instances of the scheduler process running
Output from q db f=d:
Available Space (MB): 50,012
Assigned Capacity (MB): 45,012
Maximum Extension (MB): 5,000
Maximum Reduction (MB): 10,720
Page Size (bytes): 4,096
Total Usable Pages: 11,523,072
Used Pages:
Shekhar,
Normaly that should not happen when you use passwordaccess generate.
Try to reset the client password in the admin webinterface. If this still
does not work delete the tsm.pwd file (should be locate in /tsg/tsm/security
directory as you set it up).
Then the client should ask you to
Dave,
It isn't necessary to run the service as an administrator to access the
local drive.
What usually happens is that the administrator removes the everyone
group from NTFS permissions as part of their security setup. If that is
done, you must add the system account from the local SAM back
Solaris? Which server version? How many database volumes are there defined to
TSM? How big are your database volumes? Thanks!
Luke
Remeta, Mark wrote:
I have a 40gb database on RAID5 with no performance issues.. fyi...
not sure if you consider this small, average, or heavy.
Your Cache Hit pct is too low, needs to be closer to 99%.
I would increase bufpoolsize to 131072, that's 128K and will show
as 32,768 Buffer pool pages. That should get you close or there and
eliminate the Cache Wait Pct - normally that's zero!
Expiration will be most obvious speed improvement.
Del,
Thank you, Exchange appears to be backing up right now. Strange,
though, all my other backups go to the backuppool on my TSM server's
disk, and get flushed from there to tape. The Exchange backup is going
directly to tape (I can see the tape writing while the backup is going).
Is
TSMers,
I received the following error message in my client's dsmerror.log file
last night which makes little sense - has anyone seen this before:
11/21/2002 00:11:12 Invalid subfile recognition token encountered in
transaction. Resending transaction using full file backup.
If I look at
I believe the client estimates the size of the backup and if that size
exceeds the available size of the initial pool then it tries the next pool
in the migration path and so on.
-Original Message-
From: Neil Dombrowski [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, November 20, 2002 3:42 PM
Tab,
Thanks. That's exactly what happened. I have it working now that I've given
System NTFS permissions.
Dave
At 04:36 PM 11/20/2002 -0600, you wrote:
Dave,
It isn't necessary to run the service as an administrator to access the
local drive.
What usually happens is that the
I posted this a few days ago and didn't get a hit. If anyone has anything
they could point me to or send me I would appreciate it.
Does anyone have, or know where I can get, a better technical write-up than
what is detailed by the help audit volume information the TSM server has?
I'm looking for
Hi all,
I'm moving a couple of apps - SAP/DB2 and an oracle data warehouse - from an RS6000 SP
cluster on to a couple of brand spankin' new P690s, and at the same time moving to a
new datacentre. It's just across the road but it might as well be on the moon.
1. The SP cluster has a fast
My understanding of the audit volume process is that it doesn't actually
read the data on the tape. Instead it reads the checksum information
that TSM puts on the tape periodically. This bit me once when I was
auditing a disk volume. The audit volume deleted some files out but
failed to find
Del, thank you for the information. Turns out my disk storagepool is
smaller than the database; that's why it went directly to tape. I don't
have much disk space on my TSM server, but it's probably best to send it
right to tape.
Thanks again,
Neil
On Wed, 2002-11-20 at 18:54, Del Hoobler
Hi All
I have 30 Windows 2000 Servers backing up to a Windows TSM 5.1 Server.
One of the servers when scheduled to backup begins its schedule and
backup then just sits doing nothing, it never comes back to the server
with a status and you have to cancel the session manually then it
reports back
-- Cache Hit Pct.: 90.08
David already pointed this is not good for performance.
-- Will this affect performance on the system (swap usage or paging)?
You have to verify how your operating system is tuned, how much memory is
available to applications, are there other applications, etc, etc.
Server web admin interface cannot use directly https, only the web client.
Although there is a piece of add-on code called Secure Web Administrator
Proxy. It is somewhere in your installation CDs. It allows to manage your
TSM server through some web server. Security between the browser and the
78 matches
Mail list logo
