Re: [ADSM-L] Collocation

2007-04-20 12:34:22
Subject: Re: [ADSM-L] Collocation
From: Fred Johanson <fred AT UCHICAGO DOT EDU>
Date: Fri, 20 Apr 2007 11:14:44 -0500
Just to add to Wanda's notes, if the total storage for a
particular node doesn't fill a tape, you can set up
collocation groups.  The two or more nodes will share a
tape.  If you want to move all the nodes together, MOVE NODE

---- Original message ----
>Date: Fri, 20 Apr 2007 12:10:21 -0400
>From: "Prather, Wanda" <Wanda.Prather AT JHUAPL DOT EDU>
>Subject: Re: [ADSM-L] Collocation
>1) When you set COLLOCATION ON for a storage pool, nothing
happens to existing data.  COLLOCATION just changes the rules
for TSM selecting output tapes. (actually any sequential
output volumes)
>When migration starts to a sequential pool, by default TSM
will chose a tape in the pool that is already in FILLING
status. If there isn't one, it chooses a scratch.
>If collocation is on by NODE for a storage pool, TSM will
choose a tape in the pool that is already in FILLING status
and already has data for the client node whose data is being
migrated, if there is one.  If there isn't, it chooses a
scratch.  If there isn't a scratch available, it chooses a
tape that is in FILLING status.
>Extrapolate for yourself what happens when collocation is on
by FILESPACE, by GROUP, etc.
>Same thing happens for RECLAMATION or for MOVE DATA - any
time TSM has to select a tape for OUTPUT.
>So you can turn collocation on, turn it back off, turn it on
again, etc-- All you are affecting is the choice of an output
>2) Over time, reclamation will cause the data to gradually
become more collocated, since data is written to new tapes
via the new rules.
>IF you want to force it to happen, you can run MOVE DATA or
MOVE NODEDATA commands to force existing volumes to be
rewritten, with collocation on.
>From: ADSM: Dist Stor Manager on behalf of William Kyndt
>Sent: Fri 4/20/2007 11:16 AM
>Subject: Collocation
>We currently use Tivoli Version 5.3.2 and wish to implement
>for some of our servers. I have a few questions regarding
>collocation and would appreciate any advice anyone is able
to share.
>1. When enabling collocation for a group of servers, does a
full backup
>of the servers occur to group the data on as few volumes as
>followed by incremental backups? Or do incremental backups
still occur
>as normal and are simply grouped on pre defined volumes?
>2. Is there a way to collocate data already backed up onto
>We use GroupWise as our e-mail system which comprises of
thousands of
>files and directories, when performing a full system restore
which can
>be in excess of 500 GB this can take some time because of
all the tape
>mounts taking place. We are hoping that by collocating this
data onto
>fewer tapes this will reduce the length of time taken to
complete a
>complete restore in the event of a disaster.
>Any advice surrounding collocation would be greatly
>Kind regards,
>The Miller Group Limited  Registered in Scotland - SC018135
>Miller Homes Limited  Registered in Scotland - SC255429
>Miller Construction (UK) Limited  Registered in Scotland -
>Miller Developments Limited  Registered in Scotland -
>2 Lochside View
>Edinburgh Park
>EH12 9DH
>The Information in this e-mail is confidential and for use
by the addressee(s) only. It may also be privileged. If you
are not the intended recipient please notify us immediately
on +44 (0) 870 336 5060 and delete the message from your
computer: you may not copy or forward it, or use or disclose
its contents to any other person.
>We do not accept any liability or responsibility for: (1)
changes made to this email after it was sent, or (2) viruses
transmitted through this email or any attachment.
Fred Johanson

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