What is the preferred method of shutting down the TSM server on a AIX
box when AIX is being rebooted? Is a kill command ok or is there a
kinder way to shutdown TSM without using dsmadmc?
David
To keep the archive separate, put them in a different primary/copy
storage pool.
ally do a quick check and "quiet" the TSM server personally..
Regards,
Gerald Wichmann
Senior Systems Development Engineer
Zantaz, Inc.
925.598.3099 (w)
-Original Message-
From: David E Ehresman [mailto:[EMAIL PROTECTED]]
Sent: Thursday, May 16, 2002 7:01 AM
To: [EMAIL PROTECTED]
Subject:
Make a new device class, set a max drive limit for it, and change your
copypools to use that device class.
We have it set up like you do and like John use mountlimits on the
devclasses and rigorous scheduling of the TSM events (via our MVS
scheduler, ZEKE) to minimize tape drive contention.
David
On AIX TSM 4.2.1.0, I issue a "register license file=1mgsyslan.lic
number=4" command and get back "ANR2827I Server is licensed to support
Managed System for LAN for a quantity of 1." How do I get 4 licenses
registered?
David
I am planning a TSM Server move from OS390/MVS to AIX. My largest
storage pool is about 73gig and is spread across 32 volumes of about
2.3gig each. On AIX I will have a 128gig file system so I could put
that storage pool into a single volume if that makes sense.
My question is should I spread t
In a Lan Free/Storage Agent TSM 5.1 environment, which machine issues
the tape mount command to the robot: the TSM server or the Storage
Agent?
David
>Our current schedule is:
> Weekdays 1300 till 2000 - update stgpool COPYPOOL rec=20
> Weekdays 0900 till 1300 - update stgpool TAPEPOOL rec=10
> Saturday 0800 till 2000 - update stgpool TAPEPOOL rec=10
> Sunday 0800 till 2200 - update stgpool COPYPOOL rec=20
>
Those parms say to reclaim tapes
>Upgrading from 4.1.0 to 5.1 on aix 4.3.3. If anyone knows of any
pesky
>little snakes hiding in the grass to mess up my day I would appreciate
your
>feedback.
>
There is a nasty bug in the hsm code at 5.1.0 and Aix 5.1 in 64 bit
mode. Within a minute of loading the client (including hsm) code o
>Yes, but how? There is only one scratch pool and tapes are allocated
from
>this scratch pool as needed when the primary tapepool is backed up to
the
>offsite tapeback. Is there a way to define 2 scratch pools and
allocate
>primary tapes from the Kscratch pool and offsite from the Jscratch
pool?
>We have 4 onsite pools and 2 copypools.I have defined our K tapes to
all
>onsite pools as private .
mustafa,
Are you saying that you define a given K tape to more than one storage
pool? Or are you dividing you K tapes between your defined onsite
storage pools?
David
> Are you aware of the AUDIT Library command.
>
>Generally, a tape gets in this state if it is mounted in a drive and
the TSM
>Server task crashes. The reason is TSM does not know where the tape
is
>anymore. I wish it did not work this way on the 3494 library though
>because
>the 3494 library ju
I'm pretty sure you can do this with a Servergraph schedule. Talk to
the good folks at Servergraph to find out for sure.
http://www.servergraph.com
David Ehresman
University of Louisville
Enivronment: TSM 5.1.0 running on AIX 5.1
In my /etc/profile, I have
"export DSMSERV_ACCOUNTING_DIR=/var/adm/tsm" which should cause the
dsmaccnt.log to be created in /var/adm/tsm. But it continues to be
created in the server install path, /usr/tivoli/tsm/server/bin. Any
ideas what else I nee
If the new tapes are already labeled, you just need to do the checkin.
If not, you need to do the label libvol.
David Ehresman
I am moving from an MVS server to an Aix server. When a do a "DEFINE
VOL stgpool filename FORMAT=nnn" the Aix box stops responding until
after the format is done. A monitor shows 100% i/o busy on the disk.
Is this normal Aix behavior?
TSM server = 5.1.1
Aix = 5.1
Disk = SSA Raid 5
David
I had a large backup running last night writing directly to tape. In
the morning the process to backup onsite tapes to the offsite copy
started. That process was waiting on the tape being written by the
client backup. But the client backup was stalled, i.e. it was no longer
sending data to the
Is there a parm that controls the time that tsm waits for a storage
agent to respond before issuing the "ANR8925W Drive DRIVE105 in library
3494 has not been confirmed for use by server MINERVA_SA for over 600
seconds. Drive will be reclaimed for use by others" message?
David
I have a TSM macro that issued a BACKUP STG for each of my five storage
pools. If a BACKUP STG command is issued by the macro for a storage
pool that does not have any data, I get the message "ANR2111W BACKUP
STGPOOL: No data to process." AND THE REMAINING COMMANDS IN THE MACRO
ARE NOT ISSUED.
I
What AIX (v5) command would I use to eject a given volser from a 3494
library? Where is this documented?
I am not looking for the TSM CHECKOUT command. I need to eject a
volser that is already TMS CHECKEDOUT but was reinserted to the library
in error.
David
>Will I see restore improvement on
>backupsets created with version 4.1.x.x by upgrading to 5.1 or will I
only
>see improvement on backupsets created with 5.1?
>
I am migrating from MVS TSM 4.2 to AIX TSM 5.1. As part of that
process, I am changing from backupsets to move nodedata as the means to
>But what is the improvement in restore time, which is the crucial
thing.
>
Since the restore is a normal stg pool restore, I would expect it to
take the
same time as a stg pool restore with a reduced number of tape mounts.
David
If you're using DRM, then the SET DRMDBBACKUPEXPIREDAYS sets the number
of days after which DRM will automatically delete old backups.
>>> [EMAIL PROTECTED] 08/16/02 10:48AM >>>
We are backing up our database with the following command setup in an
admin
script.
backup db devclass=ltoclass type=fu
I assume that "ANS1312E Server media mount not possible
Explanation: Server media mount not possible. The server timed
out
waiting for a mount of an offline volume." refers to mounting a
sequential acess media? If my assumption is correct, under what
conditions should a backup
>think this also:
>this volume may be full and next storage which is most probably a
tapepool
>need
>a mount point?
Which parameter controls how long a backup session will wait for a
mount of a secondary storage pool volume after the primary storage pool
fills?
The sample tsm scripts (q script) provide a number of useful select
queries.
First define all your existing scratch volumes to storage pools using
DEF VOL. Then update all your storage pools to have MAXSCRATCH=0. As
currently inuse scratch volumes go scratch, they will return to the
scratch pool but will not be reused. Periodically go back and add
recently scratched tap
When the 3494 I/O station is full of ejected tapes (with more to come),
the MVS console automatically gets a mesage "All convenience output
stations in library TRROBOT are full." Is there a way for either AIX or
a TSM server running on AIX to detect the condition so a email or page
could be sent
>Would a checkout with the "untilfull" parameter help ?
As I read the manual, untilfull requires operator intervention. The
only intervention we want by our operators is to remove the tapes and
they need something to warn them the time has come to do so. They get
an email when the move drm is d
Is there a server side message other than ANE4971I reporting zero bytes
that would alert to a LanFree node that was not using the LanFree path?
Server is TSM 5.1 running on AIX 5.1.
Given a file that has been backed up by a node, how do I find out which
tape the current copy of the file is on without doing a restore?
David
>a crude way is...for a single file...issue a restore to another
filename and
>watch
>what tape mount gets called...you can also cancel the restore once
the
>mount is requested (q sess f=d...).
>FYI
>
The restored failed without indicating which tape it thought was
inaccessable. That's why I nee
On the output from a Query Occ command, what is the difference between
the Physical Space Occupied and the Logical Space Occupied figures?
David Ehresman
>How do others manage this?
We use Servergraph/TSM. One of its many features is to send a page
when the TSM server quits responding.
David Ehresman
University of Louisville
>Another approach might be to occasionally
>(every three months maybe) do a "full" backup (by changing mode to
>"absolute" to force even unchanged files to get backed up)... this
should
>effectively "defragment" the tape pool and put all active versions on
one
>(or a couple) tape. We did this onc
Is anyone using TSM to backup Novell clusters? If so, how do you handle
the situation where a file volume fails over to another node in the
cluster?
David Ehresman
I don't believe you can do this with **scratch** tapes, but if you are
willing to predefine your tape volumes to storage groups, you can use
one tape range as your primary storage group and the other as your copy
storage group.
David
Can someone help me with a select that will list by storagepool name the
number of volumes in the storage pool and the maxscratch set for that
pool?
Thanks,
David
Anyone have data on how much overhead reconstruct=yes on a move nodedata
adds to the run time?
David
select node_name,filespace_name,description,min(archive_date) -
as "Archive Date", count(ll_name) as "Num Files" -
from archives group by node_name, -
filespace_name, description
q volhist type=backupset
Does a Reuse Delay of 1 mean "24 hours" or does it mean "after
midnight"? What process caused a tape to move from Pending to Empty
once the reuse delay time is up?
David
Aix Server patch 5.1.1.5 is apparently not generally available. Is
5.1.1.6 stable or should I go to 5.1.1.4 instead?
I am currently at Server and Storage Agent 5.1.1.0. Does the Server
and the Storage Agent have to go to 5.1.1.x at the same time or can I
take the Server to 5.1.1.x one week and
>import will usually take significantly longer then an expor
My recent experience was just the opposite. Export took days, import
took hours.
David
Is there a way to transfer a filespace from one node to another other
than restoring it on the target node and backing it up again from there
before deleting it from the source node?
David
>Depending if it is an AIX filespace or file you want to copy, you
could
>use the rcp -p /filesname nodename:/filesname command that AIX
offers.
>
How would that transfer TSM ownership?
David
Do the "under-utilized" tape volumes have data that is pre-compressed by
the client? This can cause tapes to appear to be under-utilized.
David
>Does anyone have any suggestions regarding how to calculate how much
>data
>is being handled by TSM every day ??
>
>
Get a trial of Servergraph/TSM, http://www.servergraph.com , and see if
it does what you want (and a whole lot more).
David Ehresman
A satisfied customer
> then make sure the volume has been labeled using the dsmlabel
>command, which is wierd cause i though you only use that command on
the
>drives not on tapes???
You need to dsmlabel the new tapes to write a volume serial on the
tape. dsmlabel has nothing to do with the drives.
As Paul says, this will depend as much on your data as on your hardware.
Using 3590E tape drives, yesterday we got rates of 1.6MB/s to 55.4MB/s
as reported by Servergraph/TSM on our 5 tape to tape storage pool
copies; the median was 27.4MB/s. The rates on our 8 tape storage pool
reclaims last ni
If you're running an OS390 Server, use SMS to control what data goes to
which type carts.
>>> [EMAIL PROTECTED] 03/13/02 04:20AM >>>
Hi TSM'ers ,
i have 3494 library with 3590 J and K type cartridges in it for all
sequential access storage pool,
I want to define sequential access storage pool
No, SANergy is not included in TSM v5.1. It is needed if you plan to do
LANfree to disk storage pool.
>>> [EMAIL PROTECTED] 03/14/02 05:31AM >>>
On 13-Mar-02 Cindy Bogle wrote:
> SANergy is not a prerequisite to TSM V5.1.
But my question was: "Is it included in TSM V5.1?",
and if yes, "Is i
I've not done the move yet but expect to in the next 6 months. The con
I expect has to do with media movement to and from our offsite vault.
On OS390 we have a vault management system that tells the ops what tapes
to send offsite, tells the courier which slots to put them in, which
slots to pull
Do the TSM accounting records that are written in an Aix environment
have the same layout (minus the header info) that the TSM SMF accounting
records have in MVS?
David
ill Boyer
DSS, Inc.
-Original Message-
From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED]]On Behalf
Of
David E Ehresman
Sent: Wednesday, March 20, 2002 10:13 AM
To: [EMAIL PROTECTED]
Subject: Re: Benefits of moving to platform other than OS/390
I've not done the move yet but
I currently run TSM 4.2 on OS390 2.10. I use DFSMS and the data set
name prefix set by TSM for a device class to put some storage pools on
standard length carts and some on extended length carts in a 3494
library.
I am planning for a conversion to TSM on AIX sharing the 3494 library
with MVS. O
We do a Q NODE which feeds a Q FILESPACE where we check for date of last
completed backup. This is done in REXX on OS390.
David Ehresman
University of Louisville
How would I modify the SELECT statement in the sample Q_MSG-SEV script
to only return messages from the past 24 hours?
What TSM add-on commercial programs are available for
managing/monitoring a TSM server running on AIX?
How would I modify the SELECT statement in the TSM supplied sample
Q_MSG-SEV script to only return messages from the past 24 hours?
If I put a "Q DB" and a "Q DB F=D" together in a script and run it from
a terminal (AIX), the "Q DB" output is in table format and the "Q DB
F=D" output is in list format. This is what I want. If, however, I
redirect the output to a file, "RUN QUERY > query.txt", they both come
out in table form
>Not sure what you are trying to do here.
Trying to duplicate my MVS/TSM email reports in an AIX environment :-(
David
>With all that said, I'm curious: Q DB F=D includes what is in Q DB, so
why
>do you want both?
>
Well the simple answer is I'm trying to duplicate an existing MVS/TSM
report in Aix. The better answer is that %database full stands out
better in Q DB than in QD F=D but the later has additional deta
>You could try using -outfile instead of redirection.
That did the trick. Now does anyone know why -outfile minics output to
a terminal while redirection does not? What's the difference between
them?
David
>If this is for digestion by another program, use the -comma or -tab
>switches on the dsmadmc command line.
It is not.
>Note that the WIDTH of your
>terminal window can also affect whether
>table or list format is selected.
So is there a way I can make my (unix) file the same WIDTH as my
termin
How have folks in an Aix environment handled a 3494 library with 3590
drives and a mixure of J and K (standard and extended length) carts? Do
you try to control which storage pool goes to which length cart or do
you just let the library randomly pick? If you control, how do you
control which car
On Aix after a LTO drive was replaced in a 3584, I had to rmdev the
drive and then do the cfgmgr again before TSM would recognize the drive
again. Don't know if this would help with your problem or not.
David
versions data exists - no limit
versions data deleted - no limit
retain extra versions - 30
retain only version - 30
I have a customer who tried:
incremental -verbose /ps/odstest/data
and several objects were inspected but not were backed up.
He then tried:
backup -verbose /ps/odstest/data
and that backed up the two sub directories but no files.
Finally he tried:
incremental -verbose /ps/odstest/data/
a
I there a way to get JUST the output from a select statement into a file
without the TSM version header info at the beginning and the condition
code at the end?
This is TSM 4.2 on Aix 5.1.
David
I think you have to be at TSM 5.1 to get the feature, but we use the
MOVE NODEDATA command on a regular basis to consolidate data for a node
onto a minimum number of tapes.
David
http://publib.boulder.ibm.com/tividd/td/StorageManagerforAIX5.1.html
does not work today to find the TSM Server 5.1 manuals. Where are we
suppose to find them today?
>Anyone else running 5162 and able to use the SHOW VERSION command?
I've never used the SHOW VERSION command before but on my 5.1.6.2
server, a "SHOW VERSION nodename filespacename" returns output like one
might expect from a show command.
David
Anyone backing up Groupwise postoffice boxes using TSM? If so, are you
using an open file manager that works? We're on Groupwise 6 and TSM
5.1.6.2
David
I just got back from a week off and we are having problems with backups
not finishing. The backup client is on AIX backup level 5.1.1.0 backing
up to an AIX server at TSM level 5.1.6.2. Anyone else seen problems
with this combination?
David
I opened a PMR on this issue and was told that it is documented in APAR
IC34754 with date of fix still unknown.
David
>>> [EMAIL PROTECTED] 03/03/03 02:07PM >>>
Started for us when we installed service to bring us to TSM 5.1.5.4.
(AIX
5).
Assumed it to be a 'feature'.
-Original Message-
n provide?
-Original Message-
From: David E Ehresman [mailto:[EMAIL PROTECTED]
Sent: Monday, March 03, 2003 1:25 PM
To: [EMAIL PROTECTED]
Subject: TSM 5.1.6.2 and AIX client 5.1.1.0
I just got back from a week off and we are having problems with
backups
not finishing. The backup client is o
Anyone had experiences, good or bad, with AIX backup client level
5.1.5.11?
David
Still waiting to hear from the support center. Meanwhile I found an
entry in the aix backup client level 5.1.5.11 readme that sounds like a
match. We're going to try that on one of our machines.
>>> [EMAIL PROTECTED] 03/04/03 02:51PM >>>
From: Richard Sims [mailto:[EMAIL PROTECTED]
>From someone
Is there a web document that explains the current TSM licensing and
pricing structure?
David
I've been running TSM 5.1.6.2 server on AIX in production for two weeks
in RollForward mode. Our log utilization increases steadily until we do
our daily data base backup at which time it returns to zero and the
cycle restarts.
David
>>> [EMAIL PROTECTED] 03/06/03 11:51AM >>>
Is anyone running i
>We have extended edition. But, aren't the SAN client setup as
"Library
>Clients" of the server? And if so, wouldn't this imply that Library
>Sharing is required?
>
The TSM for SAN license authorizes you to use the Storage Agent on the
client; you need one per LanFree client. You need one Libra
Is there any documentation of what messages cause which numeric return
code to the client. In particular, I'm trying to determine what return
code a "ANS1492S Invalid virtual mountpoint /ps/epmtest/psoft: File not
found." returns to the client. Is there any way to change which return
code a messa
> I saw that the cleaning tape was ejected to the bulk
>I/O area. So I removed it from the library and I placed a new one in
the
>bulk I/O area. The library picked it from there and placed it in an
empty
>cell. I though that that was enough, but apparently not. Is there a
special
>procedure for che
>One problem I am aware of at this level of the client (5.1.1.0, if I
>understand correctly) is IC34996. For AIX, this is fixed in patch
level
>5.1.5.11. You might try applying that patch to one or two machines
that
>consistently see this problem to see if that resolves it.
>
Andy,
Thanks for the
Field 17 of the TSM accounting record is "Amount of backup files, in
kilobytes, sent by the client to the server". If a file is retry-ed,
does it get counted once or twice in this field?
David
mtlib -l /dev/lmcp0 -qL
>>> [EMAIL PROTECTED] 03/10/03 03:17PM >>>
Richard,
what is the mtlib command that shows how many cleaning cycles are
available
in the library? I used to know it but nothing seems to work now.
>From: "Cook, Dwight E" <[EMAIL PROTECTED]>
>Reply-To: "ADSM: Dist Stor M
We backup the data base once a day and send it offsite. We also run the
log in rollforward mode. This means if we still have the log, we can
restore the data base and recover to the point of failure. Recovering
without the log would obviously mean we only recovered to the point of
the db backup.
>Hi David,
>
>what about AIX client 5.1.5.15 ? Have you tried it ?
>
>
René,
I have not tried the AIX 5.1.5.15. 5.1.5.11 was the highest patch
level when I started my problem analysis.
I do have 5.1.5.15 running on a handful of Window 2000/XP clients for
about a week with no problems.
David
We've been running prompted mode backups for a few months and have been
running the AIX TSM Server 5.1.6.2 for about a month and life has been
good. Last night, NONE of our prompted mode backups started. There
were no server side "ANR2561I Schedule prompter contacting VHOST3
(session 23462) to st
a
(408) 257-3037
mailto:[EMAIL PROTECTED] (change aye to a for replies)
Professional Association of Contract Employees
(P.A.C.E. -- www.pacepros.com)
-Original Message-
From: ADSM: Dist Stor Manager [mailto:[EMAIL PROTECTED] Behalf
Of
David E Ehresman
Sent: Wednesday, March 19, 2003 7:32
>The figures I quote are from the IBM Tivoli Price Estimator tool.
Is this tool publically available to TSM customers?
David
Our standard policy to retain backups for 30 days. We do that by
setting both VersionsExist and VersionDeleted to NoLimit and DaysExist
and DaysDeleted to 30.
Versions unlimited means we don't take the performance hit of
versioning during the client backup window.
David
After examining the trace I supplied, level 2 decided the problem was
probably in the TCP layer but did not have enough info to make a
definitive determination. If it happens again, they want a core dump
(Aix command kill -11 ). So if anyone else encounters this problem of
prompted mode schedules
>Where are the settings for DaysExist and DaysDeleted at?
I'm sorry. I was using the TSM/Servergraph admin gui terminology
rather than the tsm terminlogy. The tsm terms are "Retain Extra
Versions" and "Retain Only Versions" but the units are in days. These
are set on the copygroup.
David
I am experiencing the same problem on aix tsm 5.1.6.2 and support thinks
it is a tcpip problem. I would urge you to open a pmr with tivoli
support and reference pmr 14310,082
David Ehresman
University of Louisville
>>> [EMAIL PROTECTED] 03/26/03 02:12AM >>>
Hi ALL,
Iam observing a strange probl
When I get a session that does not start with a "ANR8214E Session open
with 136.165.2.36 failed due to connection refusal. " I get a session
retry every 30 seconds with the ANR8214E being issued over and over. Is
there a parm that controls that retry interval?
David
If I do a "SELECT * FROM EVENTS" I only get events for the current day.
Are the events from previous days (like those available from q event * *
begind=-1) available in another table?
David
I am moving some nodes to a new storagepool, from a non-colocated one to
a colocated one. I updated the copygroup definition to point to the new
storagepool and new backups are going there. I used move nodedata to
move the old data from its primary tape pool to the new one. My regular
backup stg
1 - 100 of 766 matches
Mail list logo
