ADSM-L

Re: [ADSM-L] SQL Query ...associated w/ management class

2010-01-29 06:24:35
Subject: Re: [ADSM-L] SQL Query ...associated w/ management class
From: James R Owen <Jim.Owen AT YALE DOT EDU>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Fri, 29 Jan 2010 06:22:51 -0500
Lindsay Morris wrote:
Right, we did some magic on the dsmc client...

But I see this question here repeatedly: "how can I list all of my backup
files?"
And the answer is always "Don't do that, or tread carefully" like Richard
just said.
...
So why are people wanting to do this?
Hey, you lurkers who think about doing this: will you speak up and say why,
please?
What are you really after, reducing wasted storage?  tuning up retention
policies?
What?
--------------------
lindsay AT tsmworks DOT com

A.Lurker responds "Yes" to both of the above, e.g.,

Upon finding DONT_DOIT policy in effect @ (inherited) NoClue TSM:

tsm: NOCLUE>q co dont_doit act*

Policy    Policy    Mgmt      Copy      Versions Versions   Retain  Retain
Domain    Set Name  Class     Group         Data     Data    Extra    Only
Name                Name      Name        Exists  Deleted Versions Version
--------- --------- --------- --------- -------- -------- -------- -------
DONT_DOIT ACTIVE    SPECIAL   STANDARD        30       30      999     999
DONT_DOIT ACTIVE    STANDARD  STANDARD         4        2       30      60

tsm: NOCLUE> [names changed, but unfortunately, this is not imaginary!]

The now responsible TSM v5.5 SysAdm wants to know (who,what,howmuch) will
be affected if we eliminate SPECIAL Mgmt Class, but would *not* dare to

select node_name, filespace_name, filespace_id, HL_name, LL_name,-
count(LL_name) as "#InActive", sum(file_size) as "NoAggr(MB)"-
from backups,contents where state='Inactive'-
and days(current_time)-days(deactivate_date)>30-
and backups.object_id=contents.object_id-
and aggregated='No'-
/* because "actual" filesize isn't always available, but it should be! */
group by node_name, filespace_name, filespace_id, HL_name, LL_name

The potential TSM v6.1.5 SysAdm wonders if, with the new DB2 basis,
this/similar selection could actually return accurate information within
the time that a reasonable person would be willing to wait?
--
Jim.Owen AT Yale DOT Edu   (w#203.432.6693, c#203.494.9201, h#203.387.3030)