Yeah, I goofed too! There is no MODE setting of CHANGEDONLY; the available
MODE settings are MODIFIED and ABSOLUTE.

Regards,

Andy

Andy Raibeck
IBM Tivoli Systems
Tivoli Storage Manager Client Development
e-mail: araibeck AT us.ibm DOT com
"The only dumb question is the one that goes unasked."
"The command line is your friend"





Steve Harris <STEVE_HARRIS AT HEALTH.QLD.GOV DOT AU>
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
07/26/2001 21:35
Please respond to "ADSM: Dist Stor Manager"


        To:     ADSM-L AT VM.MARIST DOT EDU
        cc:
        Subject:        Re: Backup Scheduling




Whoops!!

Looks like I got it backwards.  The example in the help under define
copygroup is a little confusing and could be interpreted either way (I did
check before I posted)

One reason to do a full backup is to consolidate all active files onto as
small a set of media as possible in order to facilitate a quick restore.

Steve.

>>> Andrew Raibeck <araibeck AT US.TIVOLI DOT COM> 27/07/2001 10:51:24 >>>
Steve,

I'm not sure I follow the FREQUENCY recommendation.

FREQUENCY indicates the minimum number of days that must pass before the
file is eligible for backup, regardless of whether it has changed. So if
FREQUENCY is set to 30, then the file will not be backed up until at least
30 days have transpired since the last time it was backed up, even if it
changes every day. If the file does not change at all, then it will not be
backed up via incremental backup, regardless of FREQUENCY.

The easiest way to get a full backup would be to set the management
class(es) MODE parameter to ABSOLUTE (normally it is set to CHANGEDONLY,
which causes normal incremental backup behavior). The MODE=ABSOLUTE says
to back up the file regardless of whether it has changed. (Note that this
does not change how EXCLUDE statements work; if the file is EXCLUDEd, it
will not be backed up at all).

Assuming that the only reason for running weekly or monthly "full" backups
is for the sake of getting a full backup... well, I'm not sure there's
much point to doing that. But if weekly or monthly backups are required to
be kept for a longer period of time, then this would (most likely) be a
task better suited to ARCHIVE or GENERATE BACKUPSET.

Regards,

Andy

Andy Raibeck
IBM Tivoli Systems
Tivoli Storage Manager Client Development
e-mail: araibeck AT us.ibm DOT com
"The only dumb question is the one that goes unasked."
"The command line is your friend"

Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
Sent by:        "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
To:     ADSM-L AT VM.MARIST DOT EDU
cc:
Subject:        Re: Backup Scheduling



Randall,

You can get a full backup by running a "selective" rather than
"incremental", however there is another way if you are willing to
sacrifice the exact day on which a full backup is done.

Specify a FREQUENCY parameter of say 30 days in the copygroup for these
files. Then run your daily incrementals. Any file that changes will be
backed up, and any file that hasn't been backed up in 30 days will also be
backed up again.  Provided you stagger your implementation, this should
prove as effective as a monthly full.

There is a caveat in that if you manage retention by number of versions
rather than by retention period, each selective or backup due to frequency
will push the oldest retained version off the list and it will no longer
be available. It will probably make no difference, but you should be
aware.

Regards

Steve Harris
AIX and TSM Admin
Queensland Health, Brisbane Australia



>>> Randal L Riedinger <Randal.L.Riedinger AT GRC.NASA DOT GOV> 27/07/2001
3:41:19 >>>
I'm new to TSM and I am setting up a system for the first time.  We
have around 4,000 desk top systems and 200 servers.  Most of the
desk top systems will have weekly backups of a selected directory tree.
The
rest will have monthly full backups and daily incrementals. We have
10/100 Ethernet to most of the systems and FDDI to some of the servers.
We currently have a 100 bps FDDI backbone linking the routers but we are
upgrading to 1GB Ethernet.  The TSM server has 1GB Ethernet off of
the new backbone.  We can do the small full backups on week nights and
the large ones over the week end.  The weekly backups will need to
be staggered through out the month.  If a full backup fails it needs to
be rescheduled for some number of retries with notification of the
failures.
How do I do this?  My next question is how do I set up the schedules?  I
can
visualize defining 7 weekly schedules for each day of the week and 31
monthly
schedules for each day of the month and associating the nodes to them.  Is
there
a better way to do this?  What do you have to do to get a full backup
verses an
incremental?