Re: [ADSM-L] Migration process

2007-04-10 16:06:57
Subject: Re: [ADSM-L] Migration process
From: Kelly Lipp <lipp AT STORSERVER DOT COM>
Date: Tue, 10 Apr 2007 14:06:46 -0600
Typically, a site will completely drain their cachepool before nightly
backups run.  To do this, create an admin schedule that lowers the
thresholds to highmig=0 lowmig=0.  This will fire migration processes
and continue to run them until all of the data has been migrated.  If
caching is enabled on the pool the end result will show via a q stg
operation cachepool with %Utilized 99% and %migratable 0%.  Once
migration has finished be sure to set the thresholds back to 90 and 70
or data will go directly to tape.  So either script this so that one
script does both (while waiting for migration to complete), or have
another admin schedule that runs sometime later to reset the thresholds.

An even better way is to lower to 0 0, then immediately update again to
90 0 (migration will continue since it started) and then sometime later
set low back to 70.

Update stg cachepool highmig=0 lowmig=0 migproc=4 (number of tape drives
you want to use)
Update stg cachepool highmig=90 lowmig=0

Sometime later 

Update stg cachepool highmig=90 lowmig=70

After doing this, if you are still having migration run during your
backups you need to make your pool larger by adding volumes to it.

Kelly J. Lipp
VP Manufacturing & CTO
STORServer, Inc.
485-B Elkton Drive
Colorado Springs, CO 80907
lipp AT storserver DOT com

-----Original Message-----
From: ADSM: Dist Stor Manager [mailto:ADSM-L AT VM.MARIST DOT EDU] On Behalf Of
Gregory Lynch
Sent: Tuesday, April 10, 2007 1:43 PM
Subject: [ADSM-L] Migration process

Hello All,

I have noticed that migration has been kicking off during the backup
window and is slowing down the nightly backups. We have migration
running during the day via an admin schedule for a hour, but the
diskpool never seems to get below 70%.  I checked the properties of the
diskpool and the thresholds were 90 and 70, which right off the bat I
think is too high.
What is the best practice for running migration. If it runs longer, will
it interfere with other processes like reclamation during the day?


Gregory Lynch
Lead Programmer Analyst
IT Infrastructure/Systems Administration Stony Brook University Medical
Center HSC Level 3, Room 121  ZIP 8037
Phone: 631-444-9783
Gregory.Lynch AT stonybrook DOT edu

This e- mail message, including any attachments, is for the sole use of
the intended recipient(s) and may contain confidential and privileged
Any unauthorized review, use, disclosure or distribution is prohibited.
If you are not the intended recipient, please contact the sender by
e-mail and destroy all copies of the original.