I'm starting to mess around myself.  I've not made any changes to the default 
parameters for the LIFECYCLE_PARAMETERS.  Here's the defaults as defined in the 
manual:

MIN_KB_SIZE_PER_DUPLICATION_JOB 8192
MAX_KB_SIZE_PER_DUPLICATION_JOB=25600
MAX_MINUTES_TIL_FORCE_SMALL_DUPLICATION_JOB 30

The clients that I'm using the lifecycle policy with have the following 
characteristics:
small data volumes (usually 200 MB to 500 MB each)
backups finishing all throughout the night
Writing to different media
I'm seeing some similar behavior but I do see some consolidation of 
duplications.  Some analysis in the activity monitor makes me think that the 
selection criteria is based upon when the backups finish (did they fall within 
the 30 minute window defined by MAX_MINUTES_TIL_FORCE_SMALL_DUPLICATION_JOB) 
but also if the data is on the same media.  In my batches last night, I had 
backup jobs that finished within a few minutes but did not get bundled into a 
single duplication job because they were written to different media.

Not sure if all of this information helps.  I'm still trying to get my head 
around this new feature.  It's very helpful in the environment I'm using it in 
(needing to duplicate data from my VTL to physical tape while changing 
retention level).  But, like you, want to make sure that I'm not chewing up 
extra overhead. 


> So I?m messing around with SLP?s and want to make sure the behavior I?m 
> seeing is expected. Basically what I see is that every SLP relocation is 
> broken out with one client per job. With DSSU relocation you can set the 
> STAGING_JOB_KB_LIMIT and it will group all of the clients into larger jobs 
> not to exceed that size. Here?s how I have lifecycle parameters configured:
> 
> # cat LIFECYCLE_PARAMETERS
> DUPLICATION_SESSION_INTERVAL_MINUTES 10
> MIN_KB_SIZE_PER_DUPLICATION_JOB 209715200
> MAX_KB_SIZE_PER_DUPLICATION_JOB 2147483648
> IMAGE_EXTENDED_RETRY_PERIOD_IN_HOURS 2
> MAX_MINUTES_TIL_FORCE_SMALL_DUPLICATION_JOB 30
> 
> If, for example, I backed up 10 clients of 20GB each I expect that it would 
> group the relocation of the clients into one job with a size of 200GB for 
> that job but what I get is 10 20GB jobs. It gets the job done but I have to 
> believe there is additional overhead with running that way. Is this expected 
> behavior or is something wrong? 
 :?

+----------------------------------------------------------------------
|This was sent by tom AT tjsimerson DOT org via Backup Central.
|Forward SPAM to abuse AT backupcentral DOT com.
+----------------------------------------------------------------------

_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu