Questions for sections 04 and 05
04. Frequently-asked questions on ADSM-L
04-01. Is it called ADSM, or TSM, or ITSM? What's the deal here?
04-02. What are backupsets? How can I use them?
04-03. How does TSM do full/incremental/differential backups, just like
my old backup software <fillintheblank> used to?
04-04. How do I unsubscribe to ADSM-L?
04-05. How do I do mailbox-level restores of Exchange using the Tivoli
Data Protection Agent for Exchange?
04-06. How do I force TSM to do a full backup of a client?
04-07. Where can I download the latest version of TSM/TDP?
04-08. What's the very first thing I do after TSM is delivered to me?
04-09. I'm getting message ANRXXXXX from the TSM server. What does it
mean?
04-10. I'm getting message ANSXXXXX from the TSM client. What does it
mean?
04-11. My large-scale restores are slow. How can I speed them up?
04-12. How do I back up normally open files, like database files?
04-13. What's all this about TSM and SQL select statements?
04-14. My boss wants disaster recovery procedures. What's the best way
to do it?
04-15. How do I get TSM to report problems to me?
04-16. Why does version X of TSM have this bad bug in it?
04-17. How come my copy pool tape reclamation runs so slowly?
04-18. I keep getting these "server out of license compliance"
messages. Why?
04-19. My scheduled backups fail (or are incomplete), but my manual
ones work fine. Why?
04-20. While backleveling my TSM client from 4.2.1 to 4.1.3, I get a
"downlevel" message. Why?
04-21. Why do I get an "ANR1440I All drives in use. Process being
preempted by higher priority operation" message when my
storage pool backup fails?
04-22. I've deleted all data from a tape volume, but it hasn't come
back as a scratch tape. Why?
04-23. What is this ANR9999D error message? I don't understand it.
04-24. I'm upgrading my TSM server/client from version X.X to version
Y.Y Any pitfalls?
04-25. How do I restore one client's data onto another client?
04-26. Will my new tape library work with TSM?
04-27. My Windows client backs up the same 3,000 files or so everyday.
Why?
04-28. I'm moving TSM to a new physical server. What's the best way
to do that?
04-29. How do I back up my NetWare NDS license files?
04-30. What's all this fuss about "cleanup backupgroups"?
04-31. I'm trying to include some files for backups, but it's not
working. Why?
04-32. Can I put TSM db and log volumes on raw devices?
04-33. Why is my client backup {taking so long|running so
slowly|sluggish}?
04-34. I have a tape volume that Q CONTENT says is empty,
but I can't delete the volume. Why?
04-35 I'm upgrading my TSM server from version x.x.x.x to
y.y.y.y. What's the best way to do it?
04-36 TSM is asking me to "convert" my archives? Why?
04-37 What kind/how many/what configuration should I set up for
database disks/volumes/RAIDs?
04-38 How do I move/resize my database/recoverylog volumes?
04-39 I'm moving my TSM server from operating system <BrandX>
to operating system <BrandY>. Can I just move my database
volumes from one machine to another? Why not?
04-40 My library is out of space. What's wrong with TSM?
04-41 What's the difference between a TSM database backup and a
TSM database snapshot?
04-42 How can I change the retention time for an archive I've
already created?
04-43. Boss and/or the political situation is forcing me to move my
TSM server from one operating system to another. Help!
04-44. What kind of tape drive technology should I consider for my
TSM server?
04-45 What is the "Deadly Embrace"?
04-46 What does the message 'Error 2 deleting row from table
"Expiring.Objects".' mean? Is it bad?
04-47 I've had problems using the TSAFS module on my NetWare 6.x
client. How can I make it work?
04-48 How do I back up my SharePoint Portal database?
04-49 How do I perform both full and incremental backups of my
database/mail server?
04-50 What kind of copy serialization is best?
ANSWERS to sections 04-05
04-01. Is it called ADSM, or TSM, or ITSM? What's the deal here?
The backup and restore software package we refer to today as Tivoli
Storage Manager was owned by IBM and known as Adstar Distributed Storage
Manager (ADSM). Later, Tivoli took ownership and changed the name to
Tivoli Storage Manager. The current official name of the software is
IBM/Tivoli Storage Manager (ITSM).
04-02. What are backupsets? How can I use them?
Backupsets are created from the latest (active) version of each file on
a client's system. These files are copied from data already stored in
TSM storage (disk, tape, or other media) to a single "file". Thus it is
not necessary to do an extra backup across your network to create a
backupset. Backupsets cannot be browsed like backups and archives; you
can, however, use wildcards to restore files. You can look at the
contents of a backupset, provided you know the backupset's name, from Q
BACKUPSETCONTENTS.
04-03. How does TSM do full/incremental/differential backups, just like
my old backup software <brandX> used to?
You're not using your old backup software anymore.
TSM uses a backup/restore philosophy unlike any other backup software.
Once the initial full backup of a client is finished, TSM then only
backs up files that are new or have been modified. This sort of backup
used to be called "incrementals forever", but is more accurately
described as "progressive backups". There is not a way to perform
differentials; you don't need them anymore, anyway.
04-04. How do I unsubscribe to ADSM-L?
(This question is answered twice in this FAQ because it needs to be.)
Send an email to LISTSERV AT VM.MARIST DOT EDU with a blank subject line and a
message consisting only of the line UNSUBSCRIBE ADSM-L. Do NOT try to
unsubscribe by sending email to ADSM-L AT VM.MARIST DOT EDU. All that does is
annoy the list members, and it doesn't get you off the list.
04-05. How can I do mailbox-level restores of Exchange using the Tivoli
Data Protection Agent for Exchange?
You can't, directly. Here's the scoop. You run an application called
exmerge (which comes with Exchange). Exmerge allows the Exchange
administrator to create individual mailboxes from the Exchange
information store. These individual .pst files can then be backed up
(and restored) with a regular TSM backup client.
This system of backing up individual mailboxes requires two separate
backups--one with the regular TSM client (for the .pst mailboxes) and
one with TDP for Exchange (for the Exchange directory and information
store). This could require twice as much data being moved to the TSM
server.
Notice: there is new information about the TDP for Mail (Exchange)
version 5.2.1 concerning Exchange 2003. More to come...
04-06. How do I force TSM to do a full backup of a client?
>From the command line, run
dsmc selective <drive_letter>:\* -subdir=yes
(Windows, for each drive)
dsmc selective <volume_name>:\* -subdir=yes
(NetWare, for each volume)
dsmc selective /<filespace>/* -subdir=yes
(UNIX, for each filespace)
to backup the entire drive/volume/filespace. From the GUI, press the
BACKUP button, select the desired data to backup, use the pulldown menu
to change "Incremental (complete)" to "Always backup", and press the
"Backup" button.
04-07. Where can I download the latest version of TSM/TDP?
The latest full versions of TSM backup/archive clients are available
from the ftp site
ftp://service.boulder.ibm.com/storage/tivoli-storage-management/maintena
nce. You cannot download the (required) base level of TSM server code,
but you *can* download the latest versions of the server patches at the
above ftp address. Similarly, you cannot download the (required) base
code for Tivoli Data Protection software, but you can download patches
for the base code at the above ftp address. Latest base code for the TSM
server and TDP agent are available from your Tivoli reseller.
There is a mirror of the above address at
ftp://ftp.rz.uni-karlsruhe.de/pub/mirror/ftp.software.ibm.com/storage/ti
voli-storage-management (for European users).
04-08. What's the very first thing I should do when I start using TSM?
Read the Administrator's Guide and the QuickStart manual, cover to
cover, and take notes. It's not exactly scintillating prose, but stay
with it. You'll be much happier later; I'd bet a good beer there is not
an application for which not reading the documentation can cause more
problems.
04-09. I'm getting message ANRXXXXX from the TSM server. What does it
mean?
>From the administrative command line, type
help anrxxxxx
and read the message. You can also look up the message in the TSM
Messages manual.
04-10. I'm getting message ANSXXXXX from the TSM client. What does it
mean?
>From the client command line, type
help
Navigate through the menus to find information about the message. You
can also look up the message in the TSM Messages manual.
04-11. My large-scale restores are slow. How can I speed them up?
1) Use collocated tape pools. (Just the primary pools, not the copy
pools.)
2) Use a disk-based directory management class.
3) Upgrade to TSM server 5.1 to take advantage of multithreaded restores
and MOVE NODEDATA.
4) Check your network settings to make sure that you're getting maximum
bandwidth.
5) Run a restore session for each filespace, rather than one session for
the entire node.
Number 4 above is more often than not a major culprit for this problem.
Make sure that all NICs and switch ports use optimum settings. (For
Cisco networks, hard-setting the speed (100 full duplex, 10 half duplex,
etc.) will usually get you better throughput than "auto-negotiate".
YMMV.)
Please note that numbers 1-3 are of no use once the need for a
large-scale restore comes up. They are preventative measures that must
have been in place for some time prior to the need to restore. See Rule
#1 in answer 04-14.
6) (for Windows clients) Exclude the system objects from being backed
up, run an NTBACKUP (for the system state only) as a PRESCHEDCMD, and
backup the resultant file. A restore of a Windows client in this manner
will mean that your Windows system files will restore as one large file,
rather than several thousand little ones.
7) Perform a "no query restore". The most visible benefit of no query
restore is that data starts coming back from the server sooner than it
does with "classic" restore. With classic restore, the client queries
the server for all objects that match the restore file specification.
The server sends this information to the client, then the client sorts
it so that tape mounts will be optimized. However, the time involved in
getting the information from the server, then sorting it (before any
data is actually restored), can be quite lengthy. No query restore lets
the TSM server do the work: the client sends the restore file
specification to the server, the server figures out the optimal tape
mount order, and then starts sending the restored data to the client.
The server can do this faster, and thus the time it takes to start
actually restoring data is reduced. In either case, for a given restore,
TSM will restore the files and directories in a manner that optimizes
tape mounts. If it so happens that a file is restored before it's parent
directory is restored, then the client will create the parent directory,
restore the file, then restore the parent directory when it encounters
it in the restore sequence. (Thanks to Andy R. for this one!)
04-12. How can I back up files that are normally open, like database
files?
You can use the Tivoli Data Protection agents; TDP supports most major
database formats. You can also use a third-party open file agent. You
can use other applications, like BMC's SQL-BACKTRACK, that allows TSM to
multithread most database file formats. You can temporary close the
application (or instance) that uses the file, back up the resultant cold
flat file, and then restart the application (or instance). Some database
engines allow a "dump" of data to a flat file, which can then be backed
up as a normal file. Each option has its advantages and disadvantages.
04-13. What's all this about TSM and SQL select statements?
You can write SQL select statements to query and read data from the TSM
database. (You cannot write to it with SQL.) There are some limitations
(joins are only partially supported, for example). You can get started
by running HELP SELECT from the administrative command line. You can do
some pretty neat things with SQL statements; browse the mailing list for
a plethora of examples, or look at the scripts.smp file in your server
subdirectory.
04-14. My boss wants disaster recovery procedures. What's the best way
to do it?
Answering this would take pages and pages, and much of it depends on
your resources and your business needs. TSM is a fine tool to include in
your disaster recovery planning, but setting it up properly for DR is a
non-trivial task. If you don't have the time to do the proper research
(or don't want to), contract or hire a professional who knows how.
Rule #1 of disaster recovery: Practice it before you have to.
Rule #2 of disaster recovery (the 6P rule): Prior planning prevents
pretty poor performance.
Rule #3 of disaster recovery: Document everthing! If a (literate)
stranger can't pick up your DR doc binder, turn to page 1, follow
directions, and perform the entire DR procedure accurately, you're not
documented well enought.
04-15. How do I get TSM to generate daily reports for me?
There is a TSM reporting tool from Tivoli. It's pretty spiffy. It has
both a daily report and an hourly monitor, it is highly configurable, it
can publish its reports in several formats (including popups!), and IT'S
FREE! The file is called TSMCON5220_WIN.exe (which should be installed
on your monitoring console, not the TSM server). You have to be running
TSM server version 5.2.2.0 (or higher), and it can be found at
ftp://service.boulder.ibm.com/storage/tivoli-storage-management/maintena
nce/server/v5r2/WIN/LATEST
04-16. Why does version X of TSM have this bad bug in it?
Yes, there are bugs in TSM. It is a large, complex, evolving piece of
software that has to work and play well with literally millions of
different combinations of client hardware and software; the big surprise
is that there are not *more* bugs in it. If you run across a bug (or a
perceived bug), post it to the list, or scan the APAR list on
www.ibm.com to see if it's already been reported. To get rid of the bug,
upgrade TSM to a level where the bug has been fixed.
04-17. Why does my copy pool tape reclamation runs so slowly?
If tape space reclamation tries to move files for which copies exist in
a cached diskpool, the algorithm that TSM uses for file access will pull
the file from the diskpool rather than the tapepool. This situation
causes reclamation to run *very* *slowly*.
One workaround is to turn off caching for the diskpool. Another is to
perform a periodic MOVE DATA from disk pool volumes to primary tape
pool, to flush the cache.
Note: supposedly, the most up-to-date server patches alleviate this
problem greatly. YMMV.
04-18. I keep getting these "server out of license compliance"
messages. Why?
You are using more of a particular type of client than your TSM server
is licensed to handle. Run a QUERY LICENSE to get a list of what you're
licensed for and what you're actually using. TSM runs on an honor
system; it will continue to run even if you're out of compliance. Be a
good doo-bee; purchase those extra licenses from Tivoli and look at the
HELP REG LIC response from the administrative command line (or the
Administrator's Reference).
04-19. My scheduled backups fail (or are incomplete), but my manual
ones work fine. Why?
This almost always occurs in Windows clients, and it's always a Windows
permissions problem. By default, the TSM scheduler runs as a service
with System account privileges. Also, by default, all Windows files are
read/writeable by System; however, there are a few Windows admins (and
applications) out there that remove System privileges from files. When a
scheduled backup tries to backup such files, it fails. When a manual
backup is run, the permissions of the account that is logged on are
used, rather than the System account. The real solution is reset all
files to have System read/write privileges. If that's not possible, you
need to alter the scheduler service so as to use a logon ID with
sufficient privileges to access to all files that you want backed up.
Remember that if that logon ID changes passwords, you're going to have
to update the service properties to reflect the new password.
04-20. While backleveling my TSM client from 4.2.1 to 4.1.3, I get a
"downlevel" message and no backup. Why?
When a node backs up data to a TSM server, the level of TSM client being
used is recorded as part of the node definition. When you backlevel the
client (say, from 4.2.1 to 4.1.3), TSM won't allow it to happen because
it thinks you should be using a more recent version of the client, which
may have features not available in the older version. The fix is to call
TSM support and have them step you through the process to fix the node
definition. This fix is potentially dangerous to the database, which is
why it's not distributed on the mailing list (or described here).
04-21. Why do I get an "ANR1440I All drives in use. Process being
preempted by higher priority operation" message when my
tape operation stops running?
TSM has an internal list of priorities it keeps for various
administrative events and client operations. Restores are at the top of
the priority list (obviously), client backups are at the bottom, and all
the other fall in between. When a process needs a tape drive, and
they're all busy, TSM will preempt lower-priority processes until enough
drives are freed up to fulfill the higher-priority process. For details,
look in the Administrator's Guide under "Preemption of Client or Server
Operations".
04-22. I've deleted all data from a tape volume, but it hasn't come
back as a scratch tape. Why?
Check the DELAY PERIOD FOR VOLUME REUSE parameter for the storage pool
the tape belonged to.
04-23. What is this ANR9999D error message. I don't understand it.
ANR9999D is the catchall error that (usually) indicates a problem with
low-level issues. See Richard Sim's entry on ANR9999D at
http://people.bu.edu/rbs/ADSM.QuickFacts for more details.
04-24. I'm upgrading my TSM server/client from version X.X to version
Y.Y. Any pitfalls?
Read the Quickstart manual for the version you're upgrading to. (See
subsection 3 under answer 03-02 for the location.) Also, there should be
readme files, either in the new base media or in the subdirectory where
you downloaded the maintenance level or patch. (There's a reason why the
file is called "readme".) Also, check the TSM manual for the version and
release you're working within for upgrade instructions. The procedures,
basically, is:
1) Do a fresh full db backup of the TSM server (in case you have to fall
back).
2) Follow the procedures in the Quickstart manual and/or readme files.
Also, see the answer to 04-35 below.
04-25. How do I restore one server's data onto another server?
On the target server, run
dsmc -virtualnodename=<nodename_of_source_server>
You'll have to authenticate the source server's nodename and password.
Now you can restore data onto the new server, provided that the OSs
involved are the same. In other words, you can't restore data from a
NetWare machine to a Windows machine.
04-26. Will my new tape library <brandX> work with TSM?
http://www-3.ibm.com/software/sysmgmt/products/support/IBM_TSM_Supported
_Devices_for_AIXHPSUNWIN.html shows all supported devices. You can look
for yours here. If it's not listed, it *could* work just fine; however,
Tivoli will not support it.
04-27. My Windows client backs up the same 3,000 files or so every day.
Why?
Your client is backing up its Windows system objects, most of which are
the .exe and .dll files from the \winnt\system32 folder. Unless you
exclude system objects, they're going to get backed up every time you
back up the Windows box. Warning: if you exclude system objects, the
Windows registry will not be saved. See item #6 in answer 04-11 for an
alternative to backing up system objects.
04-28. I'm moving TSM to a new physical server. What's the best way
to do that?
There are several ways to do it. Here is one that many people get good
results with:
1) Install TSM on the new box/get the library set up/etc.
2) Point all the clients at the IP address for the new server by use of
the client option file.
3) Use the new server for backups, and the old server for restores only.
(You'll have to change the IP address of the TSM server in the option
file for any client that needs a restore done. Don't forget to change it
back when you're done.)
4) Let the data on the old server expire away for a period of time. When
the old data expires down to a point to where you can stand it, use
EXPORT NODE on a server-to-server connection to move the data across to
the new box.
Believe me, this is a lot easier than trying to export all of your old
server's data to the new box all at once. A lot.
04-29. How do I back up my NetWare NDS license files?
You can't, and you wouldn't want to anyway, because in the case of a
restore you'd have to reinstall the app that placed the license file in
NDS in the first place. You can exclude the license files by using
EXCLUDE "NDS:*License ID*"
in your NetWare client's exclusion list; this allows you to avoid the
annoying "can't back this file up" messages you'll get otherwise.
04-30. What's all this fuss about "cleanup backupgroups"?
If you used TSM server versions 4.1.x or 4.2.1.x, you most likely
encountered a bug in the server code that caused the system objects in
your Windows clients' backups to not expire properly. An upgrade to
4.2.3.x or 5.1.5.x fixes the bug, but you have to run
cleanup backupgroups
to rid yourself of unwanted versions of the system objects. This subject
has been talked to death in the mailing list; go to
http://search.adsm.org to review the details.
04-31. I'm trying to include some files for backups, but it's not
working. Why?
If you have an EXCLUDE.DIR line in your include/excluse list, it will
override all INCLUDE lines, no matter what the line placement is.
04-32. Can I put TSM db and log volumes on raw devices?
Yes. On AIX, there may be some performance enhancement, due to the fact
that JFS won't log the changes.
04-33. Why is my client backup {taking so long|running so
slowly|sluggish}?
See "Backup taking too long" from
http://people.bu.edu/rbs/ADSM.QuickFacts. [Thanks to Richard Sims for
this one.]
04-34. I have a tape volume that Q CONTENT says is empty,
but I can't delete the volume. Why?
You've got a discrepency between the TSM database and what is actually
stored on the tape; the volume is most likely empty, but the database
doesn't realize it. Run an AUDIT VOLUME <volume_name> fix=yes; that'll
take care of it.
04-35 I'm upgrading my TSM server from version x.x.x.x to
y.y.y.y. What's the best way to do it?
After taking a substantial poll of postings from the last six months, it
doesn't appear that there *is* any best way to do it. When you go from
one major version to the next (i.e., from 4.2.2 to 5.1.0), you'd best
have base media with you, since license file formats generally change
from one set of base media to the next higher version. We have reports
of success (and failure) for upgrades from one patchlevel to the next
patchlevel, and from as far back as version 3.1.x to version 5.1.6.x. As
always, read the documentation accompanying the upgrade code *first*.
(See the answer to 04-24 above.)
04-36 TSM is asking me to "convert" my archives? Why?
>From the HELP CONVERT ARCHIVE message:
"Use this command to improve the efficiency of a command line or API
client query of archive files and directories using the description
option. Previously, an API client could not perform an efficient query
at all and a Version 3.1 or later command line client could perform such
a query only if the node had signed onto the server from a GUI at least
once. Perform this conversion only once for each node."
04-37 What kind/how many/what configuration should I set up for
database disks/volumes/RAIDs?
There are no conclusive "best" ways to set up the TSM db and its
volumes, other than the common-sense ones:
1) The more spindles, the better.
2) The faster the disk RPM, the better.
3) The larger the cache (for higher-end disk boxes), the better.
4) RAIDs are a wash in the long run, because they all have a penalty on
either reads or writes. Some RAID configurations, however, do give a
measure of fault tolerance in case of hardware failure.
04-38 How do I move/resize my database/recovery log?
There are several steps involved in this: 1) define dbvol/logvol for
your new volumes; 2) extend the db/log by the amount of volumes you've
added; 3) run Q LOGV/DBV F=D and document how large your old logv/dbv
volumes are; 4) reduce db/log by that amount; and 5) delete the unwanted
dbv/logv by running DELETE LOGV/DBVOL.
Notice that the actual db/log volumes files still exist, even though TSM
does not use them anymore. You will need to manually delete the volumes
at the OS level.
04-39 I'm moving my TSM server from operating system <BrandX>
to operating system <BrandY>. Can I just move my
database volumes from one machine to another? Why
not?
This will not work, and it is certainly not supported. The reasons lie
partially with the bigendian/littleendian issues that different
operating systems use to structure their filesystems. The only supported
way to move your TSM server between dissimilar operating systems is to
run an export/import process, which could require TSM to be down a long
time dending upon the amount of data being moved. If you're getting a
new tape library in addition to a new server, then your best bet is to
put the new server online and perform all new operations on it, allowing
the data on the old server to expire away. After several months, the
remaining data on the old server
04-40 My library is out of space. What's wrong with TSM?
(From Wanda Prather, with a tip 'o the hat) I'm not sure this is an
issue of TSM design. If your libraries are out of capacity in terms of
SLOTS, rather than throughput, you just have "too much" data. That
either means you are
1) not compressing the data as much as you can
2) backing up things you don't need to
3) keeping data longer/more copies than you need to
4) really in need of additional library space
5) not running expiration
6) not reclaiming often enough
For #1, it's a matter of checking to make sure that your drives do have
compression turned on. If you can't compress at the drive level, turn
it on at the client level. For #2 through 4, I don't know any
magic/automatic way of figuring it out. Here's what I do:
dsmadmc -id=xxxxx -password=yyyyyyy -commadelimited "select
CURRENT_DATE as DATE,'SPACEAUDIT',node_name as node,
backup_mb,backup_copy_mb,archive_mb,archive_copy_mb from auditocc"
Suck that into a spreadsheet and look to see which clients are taking up
the most space on the server side. Then go look in detail at the
management classes and exclude lists associated with the "hoggish"
clients, and see what you can find out about the copies they are
keeping.
Are you keeping copies of EVERYTHING on the client for a zillion
versions, rather than just the important data files?
(for Win2K) Are you keeping more copies of the SYSTEM OBJECT than would
likely be used?
Look at their dsmsched.log files and see what is actually being backed
up.
Be suspicious of TDP clients not deleteing copies they are supposed to.
(For example, if they are supposedly keeping 10 versions of a 10 GB
database, but the SELECT shows 500 GB on the server, there's something
wrong.)
If it's user/group space, are there lots of .mp3 files? (exclude 'em
with a clientoptionset)
Make sure you aren't backing up TEMP directories
etc..
I run the query monthly and save the data so that I can compare from one
month to the next. That tells me which clients are GROWING the fastest.
Those are the ones to attack. With luck, you will find some things that
you can do that will extend your library life a while. Maybe not. But
at least you will be able to
tell your management WHY you are running out of space.
Check your dsmaccnt.log accounting log. Pull the backup/archive data
from it, create a daily summary of how much data was backed up/archived,
and graph it using your favorite spreadsheet. See if there is an upward
trend...
04-41 What's the difference between a TSM database backup and a
TSM database snapshot?
(from Mark Rodriguez) DBSNAPSHOT - should be used for your DR backups of
the DB. They should be to a tape and taken offsite with your DR tapes.
DBSNAPSHOT is effectively a full backup of the DB, but it does not
involve the LOG in any way. Therefore, it can only be used for a PIT
(point in time) restore. Also, please note that a DBSNAPSHOT does not
reset the LOG like a DBB (full or incremental) does if the LOG is set to
ROLLFORWARD. This makes it perfect for a DR type restore since in a
disaster you would expect that the LOG would be lost. Please note if
you are using DRM then you should update your scheduled Prepare command
to include "Prepare Source=DBSnapshot" so that it manages the correct
tapes for you.
DBBACKUP (full and incremental) - should be kept onsite for rapid DB
recovery including restore to the most recent time using a ROLLFORWARD
log. I prefer to do these DBB to a device of type FILE. This allows
for much faster backups and restores. Also, if you use DEFine
DBBackuptrigger you can set your triggered backups to go to the same
device of type FILE which means in a triggered event your DBB will take
place much faster and a rapidly growing DB would be less likely to
trigger the DB spacetrigger.
04-42 How can I change the retention time for an archive I've
already created?
Just change the retention rule for the management class under which the
archive was originally created. All files are updated with the new rule.
04-43. Boss and/or the political situation is forcing me to move my
TSM server from one operating system to another. Help!
Your best bet is to create the new TSM server, set both TSM servers to
share the tape library (or set up a new library if you're getting one),
start performing all backups with the new TSM server, and allow the data
to slowly expire away on the old TSM server. You will eventually have to
use a node export to move data to the new server, but it will be a lot
less painful if you allow several months of expiration first.
If you don't have a second library (or your single library is not
shareable), you will have to do an export from tape (from the old
library) to disk (on the new library), and you're going to have to have
enough disk to contain the entire occupancy of the library. *Warning*:
the export will be very slow. There are other alternatives, but most of
them are pretty distateful.
The bottom line is: convince Boss to let you stay right where you are.
04-44. What kind of tape drive technology should I consider for my
TSM server?
(A tip of the hat to Wanda Prather for this one.)
Rember that I/O, I/O, it's all about I/O - don't skimp on fast disk for
the server, enough buses to your tape drives. When in doubt, assume you
will have the same performance issues as on your mainframe - the more
paths the better.
Don't sweat it too much for the initial configuration - as one responder
said, just make sure you have enough expandability to add adapters &
memory in the future, as you need them.
WHATEVER you do, don't skimp on tape drive quality- TSM beats them to
death. Many of the types of drives you see people using in network
environments are cheap for a reason - they just aren't as durable under
load as the types of drives you are used to in a mainframe environment.
Based on personal experience I would not consider ANYTHING but LTO2,
IBM3590, or STK9840/9940 in terms of speed/reliability. All those work
great. (But because you expect rapid storage growth, that probably rules
on the 9840, maybe the 9940 also.)
04-45 What is the "Deadly Embrace"?
When the TSM server is in the middle of processing a large file,
cancelling the process involved requires waiting until the file is dealt
with. This can be mean a long wait (on a storage pool backup utilizing
slow tape technology, for instance) or an interminable wait (a tape
drive starts spitting out write errors). The only way to "force"
cancellation of a such a process is to stop and restart the TSM
service--a cure that may be worse than the disease.
04-46 What does the message 'Error 2 deleting row from table
"Expiring.Objects".' mean? Is it bad?
It means that the current version of the TSM server you're running has a
bug in it (specifically, APAR PQ87485). The solution is upgrade your TSM
server to version 5.1.9.3 (if you're on 5.1.x) or 5.2.3 (if you're on
5.2.x).
04-47 I've had problems using the TSAFS module on my NetWare 6.x
client. How can I make it work?
[from Troy Frank] "I've got mixes of TSA500, TSA600, and TSAFS depending
on the OS version. The only one I've ever had any problem with is
TSAFS, and even that has cleared up in the lastest couple service packs.
If TSAFS is giving you trouble, the key is to limit how much cache
memory that it is allowed to allocate. This configuration can be done
via commandline switches is smsstart.ncf, or through iManager under the
"SMS" heading. The setting is called "Cache Memory Threshold". I would
suggest starting with a value of 10, and moving progressively down to 1,
until it stops causing problems. The higher you can keep it, the better
performance will be."
04-48 How do I back up my SharePoint Portal database?
There is a utility built into SharePoint that facilitates a cold dump of
the SharePoint database. Refer to Microsoft article
http://support.microsoft.com/default.aspx?scid=kb;en-us;281413 for more
details.
04-49 How do I schedule full/differential/incremental backups of my
database/mail servers that have TSM clients for those
application?
You will need to set up a client schedule on the TSM server for each
type of backup you intend to do. In each schedule definition, the OBJECT
parameter should contain the directory tree and filename of the batch
file on the TSM client that runs the command-line version of the backup
type you want to run. The client schedule should then be associated with
the nodename(s) that points at the particular TSM client you wish to use
for the backup. That nodename should, in turn, be associated with the
proper dsm.opt client option file (the one belonging to the specialized
TSM client, not the standard b/a client).
Consult the proper TSM client manual for the syntax of the command you
wish to run; the client manuals also contain more detailed instructions
on how to set up multiple types of backups.
04-50 What kind of copy serialization is best?
(From Andy Raibeck) STATIC tries to back the file up once. If it can't
capture an unchanged version, then the file is skipped.
SHRSTATIC will try up to 4 times to back up a changing file. Though it
will skip the file if it ultimately cannot get an unchanged version, at
leat you get several tries. The downside is that when the file is
retried, the current transaction (which can include other files as well)
has to be resent, and thus represents an increase in backup run time.
In general, having a good backup version takes precedence over
performance, and therefore SHRSTATIC is what I would recommend over
STATIC. In general, use of STATIC or *DYNAMIC methods should be on an
exception basis only, when the trade-offs between serialization, value
of the data, risk of fuzzy backp or no backup, etc., are fully
understood.
05-01. I have a Tivoli ID and an IBM.com registered ID. Which one do I
use for problem submission?
If your company has NOT been migrated to a Passport Advantage agreement,
you can use your Tivoli ID to logon to ESR. To reach entitled data on
IBM.com, you would continue to use your IBM.com ID. Once you are
migrated to a Passport Advantage agreement and customer number, your
IBM.com ID will work for both areas.
05-02. The top of every web page has a search bar in a blue field and
4 links in black. When these are used, I leave the Tivoli
content areas, why?
The top blue field area, is used exclusively for searching the web page
tags across the entire IBM web site. The links in black are references
for all IBM segments, not just software products. When using only the
software support function, use only the links below the top 10% of the
page or the black and blue fields.
05-03 Does Tivoli have a Support home page and where is it?
Yes, We have created a Support Home page, on IBM.com. It can be found at
http://www-3.ibm.com/software/sysmgmt/products/support/. Many of the
features available from Tivoli.com/support, are selectable in the left
and right navigation areas. For example, the 'Library' link, replaces
the Tivoli Information Center.
05-04 How do I use the Search capabilities on IBM.com?
For an overview of the Search system, you can reference the Search site
tour at:
http://www-3.ibm.com/software/support/viewlet/swsearch/sw_search1_viewle
t_swf.html
Once a search has been run, a count of the records is found listed in
the dark blue bar, just above the records found. To further isolate to
the information needed, additional keywords, using the categories, or
types of support features from the search results page, will reduce the
list of records found.
05-05 Where can I find Product Downloads?
On each product page under the self help section, a link will be found
if downloads for that product exist. Once this is selected, further
isolation can be done using the search results page features. Downloads
are also searchable from the IBM Software Support page, using the center
column "Downloads" link under the Support Search bar. Once the results
list for downloads is displayed, you can use isolation features that
will limit hits to only a certain Operating System, or only to the
application at a certain version / release level.
05-06 Are FAQ's documented online?
A: We have received a number of questions about 'How to find' or 'How to
use' content or features on the new support location. To help provide a
available reference for these questions, Tivoli is maintaining a Support
Site FAQ page . This link is
http://www-3.ibm.com/software/sysmgmt/products/support/Support_Site_FAQ.
html
--
Mark Stapleton (stapleton AT berbee DOT com)
Berbee Information Networks
Office 262.521.5627
|
