ADSM-L

Re: [ADSM-L] Collocation by group

2007-06-22 15:36:32
Subject: Re: [ADSM-L] Collocation by group
From: Richard Mochnaczewski <Richard.Mochnaczewski AT STANDARDLIFE DOT CA>
To: ADSM-L AT VM.MARIST DOT EDU
Date: Fri, 22 Jun 2007 15:32:25 -0400
Actually, on my tape pools I had to enable collocate by group instead of node. 

I don't see the point in doing a move nodedata. Why would I want to collocate 
all my data on that node ? If I run a selective on all filesystems, I would 
have a full image of what the syestem looked like today. If I had a DR 
situation in a week , then 
I know that most of my data is mostly on one tape. If I do a move nodedata and 
move all the data, I would have a lot more tapes to deal with. I would rather 
let retnetion/expiration come into play and drop the older files and keep my 
collocation "clean".

Rich

-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU]On Behalf Of
Nick Laflamme
Sent: Friday, June 22, 2007 3:13 PM
To: ADSM-L AT VM.MARIST DOT EDU
Subject: Re: [ADSM-L] Collocation by group


Richard Mochnaczewski wrote:
> Hi *,
>
> I am trying to setup collocation by group ( running TSM 5.3.3.0 on AIX 5.2 
> ML9 ). I created the collocation group and added a new node to to it which 
> has been backed up in TSM before but not using collocation. The first backup 
> as a collocated group menmber was a "Full" incremental . I checked my 
> collocation disk pool and the data was in there .Then I added another node to 
> the group which had been backed up to TSM and was collocated by node. I added 
> it to the group, ran another "full" incremental, checked if the data as in 
> the collocation pool  and then ran a backup stgpool to created my offsite 
> tape. When I check the contents of the tape, only the first node's data is on 
> it. Did I miss something ? I checked the definition of my collocgroup and 
> both nodes are present and accounted.
>
> Rich
>

I'm going through a similar process right now. I think you want a MOVE
NODEDATA command to move the second node's data into the collocated
storage pool.

As it is, when the client runs what you call a "full incremental," the
client sees that most of the data is already on the TSM server -- and it
doesn't care that it's not in the storage pool you want it in. (Why
should a client care about storage pools, right?).

Hope this helps,
Nick

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