How to guarantee that OST Media Server/client will use its own resources for its backups

[jbrayton]

Based on our observations when MUST_USE_LOCAL_DRIVE is enabled:

It changes the behavior of the OST Media Server selection by telling netbackup to use the Media Server resources on the client to perform the backup (whether disk or tape). 

It does not always guarantee that if the client is also an OST Media Server/data mover then will be used to handle its own backups.  

Is there a way to do the following

How to guarantee that OST Media Server/client will use its own resources for its backups when this setting is enabled.  It would greatly simplify management with multi-node database backups (e.g. Exadata, Oracle/RAC) where each node is a data mover and a single backup policy/Stu/SLP is utilized.

[RonCaplinger]

I don't think OST has anything to do with the problem you see.  However, I've been wrong before!

I found this note from Marianne, you might be having the same issue and it might be "by design":

https://www-secure.symantec.com/connect/forums/mustuselocaldrive-and-stug-70

There is also another directive "DONT_USE_SLAVE" that may help.  I'm not sure if this is still valid in 7.x, nor do I know if it affects anything regarding drives that aren't down:

http://www.symantec.com/business/support/index?page=content&id=TECH19831

[jbrayton]

No we are not using storage unit gorups

.  Below is the STU used with the Oracle/RAC backups.

Label:                dl980node-ny3dd67001-ost
Storage Unit Type:    Disk
Media Subtype:        DiskPool (6)
Host Connection:      dl980node1.dev.us.ml.com
Host Connection:      dl980node2.dev.us.ml.com
Concurrent Jobs:      32
On Demand Only:       yes
Robot Type:           (not robotic)
Max Fragment Size:    524288
Max MPX:              1
Block Sharing:        no
File System Export:   no
Ok On Root:           no
Disk Pool:            ny3dd67001-data-dp