ADSM-L

Re: scheduling questions

2005-02-28 14:22:30
Subject: Re: scheduling questions
From: "Mark D. Rodriguez" <mark AT MDRCONSULT DOT COM>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Mon, 28 Feb 2005 13:21:20 -0600
I think you got the idea now.  What you propose would work just fine.
Now the only thing you have to consider is the type of image backup you
will do.  Please read the manual on this since you will have a few
options and it will really depend on your particular environment.

--
Regards,
Mark D. Rodriguez
President MDR Consulting, Inc.

===============================================================================
MDR Consulting
The very best in Technical Training and Consulting.
IBM Advanced Business Partner
SAIR Linux and GNU Authorized Center for Education
IBM Certified Advanced Technical Expert, CATE
AIX Support and Performance Tuning, RS6000 SP, TSM/ADSM and Linux
Red Hat Certified Engineer, RHCE
===============================================================================



David Soucy wrote:

Mark,

I guess I'm just wondering how I could have daily, weekly, and monthly
backups automated under a single node name.  What I'm gathering is that
you would have to schedule dailys (M-F) a weekly image backup (Saturday)
and a monthly archive (Sunday).  You can have 3 different data retention
policys this way:

(dsm.opt)

Include E:\...\* DAILY_MGMT_CLASS
Include.Image E: WEEKLY_MGMT_CLASS
Include.Archive E:\...\* MONTHLY_MGMT_CLASS

This would be the correct (or makes the most sense) way to handle all
three of my schedules, and three different data retention policies,
right?

-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of
Mark D. Rodriguez
Sent: Monday, February 28, 2005 12:28 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: scheduling questions

David,

I don't see why you would need multiple nodes on a single system.  You
can simply run a separate schedule that does image backup.  Then you can
use include.image statements to control the management class that the
image is bound to.

I think that answers your question, but if mot repost and I will see if
I can expand my answer.

--
Regards,
Mark D. Rodriguez
President MDR Consulting, Inc.

========================================================================
=======
MDR Consulting
The very best in Technical Training and Consulting.
IBM Advanced Business Partner
SAIR Linux and GNU Authorized Center for Education
IBM Certified Advanced Technical Expert, CATE
AIX Support and Performance Tuning, RS6000 SP, TSM/ADSM and Linux
Red Hat Certified Engineer, RHCE
========================================================================
=======



David Soucy wrote:



Sounds good.  I have just been getting into experimenting with image
backups.  My only problem is how do I schedule these so they're not a
manual process?  The reason I thought I might need multiple nodenames


in


TSM is to get around using a single dsm.opt specifying what management
class to use for the filesystem I'm backing up.  I didn't think you
could specify that when setting up an automated schedule on the TSM
server...?

Is this wrong?  Or do I need to set up the schedules to run a 'command'
instead of 'incremental', 'archive' or 'image backup' when I set them


up


through the TSM server GUI?

-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf


Of


Mark D. Rodriguez
Sent: Friday, February 25, 2005 12:11 AM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: scheduling questions

David,

Mark's suggestions below will certainly work, but as in all things TSM
there are many ways to accomplish what you need.  I have a customer


with


a similar requirements.  He needs daily incremental, weekly snapshot
(full image) kept for 3 months, quarterly snapshot (again full) kept


for


a year, and as if that isn't enough a yearly snapshot kept forever!


For


the quarterlys and year end we are doing manual archives.  The reason


we


don't schedule them is more related to their internal processing, i.e.
waiting for the books to close for the quarter or the year.  But for


the


weekly snapshot we are using image backups.  I am not sure what your
client types are so I would recommend you look at the different


features


of image backups since they vary between windows, Linux and other
Unix's.  We have both windows and AIX.  There are some very good
advantages to image backups over archives.  The backup time vs. the
archive time is much shorter.  The image process will also improve your
restore process in the event you have to restore the entire file
system.  Furthermore, the image backup will require fewer resources


then


the archive will, since the image is just one object in the TSM DB as
opposed to however many files there are in your file system.

So I recommend that you at least look at image backups and see if it
will work in you environment..

--
Regards,
Mark D. Rodriguez
President MDR Consulting, Inc.

=======================================================================


=


=======
MDR Consulting
The very best in Technical Training and Consulting.
IBM Advanced Business Partner
SAIR Linux and GNU Authorized Center for Education
IBM Certified Advanced Technical Expert, CATE
AIX Support and Performance Tuning, RS6000 SP, TSM/ADSM and Linux
Red Hat Certified Engineer, RHCE
=======================================================================


=


=======



Stapleton, Mark wrote:





From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On
Behalf Of David Soucy






I was wondering if some of you could share your scheduling


strategies.


For most nodes, I just use a standard daily incremental
schedule.  The problem is that we have to keep all of our data
for year.  So, for some servers I am keeping 365 versions of a file.

I am trying to figure out a way to keep up to 14 days of
daily backups, then maybe I keep a weekly backup for 6 months
and I keep a monthly backup for a year.  I'm pretty sure I
need at least 3 schedules to do this, but do I need to
register 3 node names too?






Piece of cake. Use archives for your long-term backups. Create a


weekly


archive client schedule with a six-month retention, and a monthly
archive schedule with a year's retention. You'll end up with one
nodename and three schedules, all associated with the one nodename.

--
Mark Stapleton (stapleton AT berbee DOT com)
Berbee Information Networks
Office 262.521.5627














<Prev in Thread] Current Thread [Next in Thread>