Netbackup Staging not enough time

[Olssa]

Hi.

I have a Master/media server with two STU that i stage to tape (Tape library with two lto 4 drives)

The problem is that theres not enough time for the staging procsses to finish before the backup jobs starts.

Staging begins at 06am and backup jobbs start at 06pm.  The tapes work at full speed.

Is there any way to stop the staging before the jobs starts?

When i manually cancled a duplication job, is the images that got duplicated readable?

I dont have license for SLP.

Better solutions is welcome, I diddnt set this up, i just got it from a former colegue. =)

[Nicolai]

You can define windows where stagning is allowed to take place. But duplications already on-going will not get canceled.  To define a windows go into Storage Unit - Slect Storage Unit and click "Staging Schedule"

Images staged (processed) in a batch will be readable on, the image currently undergoing duplication will not be readable. 

DOCUMENTATION: Description of Symantec NetBackup Disk Staging Relocation Behavior

http://www.symantec.com/docs/TECH44719

PS: Have you tuned NUMBER_DATA_BUFFERS and SIZE_DATA_BUFFERS, there is great performance improvement to find.

Where are the values for SIZE_DATA_BUFFERS, NUMBER_DATA_BUFFERS and NET_BUFFER_SZ located on Windows NetBackup Servers
http://www.symantec.com/docs/TECH19206

[Marianne]

Please let us know what you perceive to be "full speed"  (The tapes work at full speed.)

What speed are you seeing (Kb/sec)? If drives ARE working at full speed, there is nothing you can do. If they are NOT writing at sustained speed of more than 100 Mb/sec, they are not writing at 'full speed'.

Please create bptm log to see what is happening on the tape creation level. We need to see if bptm write process is waiting for full or empty buffers.

[mph999]

I agree with Marianne, if your drives are running well then it looks like you have an overloaded system, there are only so many hours in a day ... I personally believe a system should never run 24/7, as, of there is unplanned downtime, you are unable to catch up. Martin

[Mark_Solutions]

There should not be an issue in staging running while backups are running - slp's will do exactly that If you want to be efficient set your staging window to start 1 hour after your backups begin and with a frequency of 1 hour - leave it open until mid morning - that is as close to an slp as you will get As others have suggested check your buffer settings also

[Olssa]

I'll dubble check the speed in perfmon tommorow and come back!

Do you have a guide on how to create the bptm log, i'm, quite new to this.

[Marianne]

Create bptm folder under ...\veritas\netbackup\logs (Windows)  or /usr/openv/netbackup/logs (Unix).

No need to restart NBU. Logs will be created in bptm folder next time bptm process is called.