Video Screencast Help

optimized duplication

Created: 07 Jan 2013 • Updated: 08 Jan 2013 | 1 comment


our environment: NBU on Windows 2003, Advanced Disk Pools, 2x Quantum DXI (VTL and OST).

I want to optimize our duplictions. They are managed by SLPs. We have many small duplication-jobs because we reach the entry


So I want to know if we can condense duplications of many media servers.

Our Support refer me to:

but I don´t know the impacts of the different settings.

Has anyone tested the settings?



Comments 1 CommentJump to latest comment

loori's picture

AFAIK you cannot bundle duplication jobs created by SLP including more than one media server because you can't bundle jobs concerning more than one DSTU. So the number of duplication jobs increses with the number of DSTU. If enabled in in the lifecycle parameter file (DUPLICATION_GROUP_CRITERIA 1) netbackup creates duplication batches by grouping by DSTU, data classification, retention period and priority, with a size limit you can set in the parameter file. See If you can accept the delay you can increase MAX_MINUTES_TIL_FORCE_SMALL_DUPLICATION_JOB and use MIN_GB_SIZE_PER_DUPLICATION_JOB and MAX_GB_SIZE_PER_DUPLICATION_JOB to optimize the jobs. I strongly advise to monitor the age of the oldest not duplicated image. The provided reports in OPS-Center are useless.

I can't see a connection to the provided link. The problem described there is about the media server which duplicates the backup, if it's the same as the one which did the initial backup and the duplication is therefore done "locally", or the duplication is done over the network. In our case it doesnt't matter because we use a 10GBE backbone with two NIC in active/active bond. That decreases problems with unneccessary waiting for free drives. When you want to accept that you can force a "local"  duplication by setting that parameter to "required".