What I do is I run around 4 (this is of course totally dependent on the
configuration of your TSM environment) identify proc's during backup and
morning housekeeping processes as not to increase the time it takes to get
my copypools up to speed to much (or slow down backups).
As soon as my offsite database backup is done I increase the amount of
identify processes to around 12 (this server has 18 lun's for the filepool,
each on
his own raid array).
This works really well because I don't notice the 4 processes during the
backup but they do a nice amount of work without any noticeable change in
backup windows so I can ingest more none-deduped data per day.
The identify process does not seem to care about the location of the
volumes it works on, it seems to think it's totally fine to work on 5
volumes in one mountpoint and 0 on another to start with, this seems (to
me) like a small point for improvement because having 1 or maybe 2 of these
per mountpoint and having the system balance that out a little better would
mean we could get even more none-deduped data in per 24 hours.




On Wed, Sep 4, 2013 at 10:12 PM, Prather, Wanda <Wanda.Prather AT icfi DOT 
com>wrote:

> Yes, you can run them at the same time.  Normally I leave Identify
> Duplicates running all the time, and leave the reclaim for the dedup pool
> set to 15% so that TSM will fire the reclaims as soon as the Identify
> Duplicates finishes with a volume in the pool.
>
> That being said, there is of course a performance impact on the database
> any time EXPIRE INVENTORY is running, so that could impact you if
> performance is already an issue.
>
> Also, I believe the server parm defaults to DEDUPREQUIRESBACKUP YES.
> Which means you also need to have your BACKUP STGPOOL completed before the
> identify duplicates will do you any good.
>
> W
>
>
> -----Original Message-----
> From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf 
> Of
> Robert Ouzen
> Sent: Wednesday, September 04, 2013 4:08 AM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: [ADSM-L] Identifi duplicates question
>
> Hello
>
> I need badly to release space of one of my STG configure with Dedup of
> Tsm6  (deduplication on the server side only)
>
> I run identify  duplicates , found a lot of duplicate extents ... my
> question is If I can in the same time to run an expire inventory and after
> a reclaim stg ?
>
> My TSM server version is:  6.3.4
>
> Regards Robert
>