Netbackup 7.1 - High&Low Watermark best practice

Hello guys,

Problem is that sometimes my B2D space ran out of space and the backup will abort. So the question is, if the Settings of my Storage Units look like they should look to work properly:

  • Max concurrent Jobs: 4
  • Reduce Fragment Size to: 4000 MB
  • High water Mark: 90%
  • Low Water Mark: 0%
  • Enable Temporary Staging Area: yes

My Setup looks like this: 5x B2D each 2 TB for the daily and the weekly setup, 2x B2D each 2 TB for the monthly setup. NetBackup 7.1

For me it seems like the diskspace is running low, before the Staging to the tapes is starting. Is there any changes I can make on the High Water Mark to start the job from the B2D to the tapes earlier or any other improvements, which make sense?

Thanks for a short reply

regards

Marc

1 Solution

Accepted Solutions
Accepted Solution!

"Is there any changes I can

"Is there any changes I can make on the High Water Mark to start the job from the B2D to the tapes earlier..."

High water mark does not initiate duplication to tape - the schedule in your DSSU config controls this. 
Check duplication schedule - ensure it runs often enough (not just once a day).

High water mark initiate cleanup of backups that were already successfully duplicated.

A couple of DSSU TNs that describes staging and cleanup:

http://www.symantec.com/docs/TECH66149 

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

View solution in original post

4 Replies
Accepted Solution!

"Is there any changes I can

"Is there any changes I can make on the High Water Mark to start the job from the B2D to the tapes earlier..."

High water mark does not initiate duplication to tape - the schedule in your DSSU config controls this. 
Check duplication schedule - ensure it runs often enough (not just once a day).

High water mark initiate cleanup of backups that were already successfully duplicated.

A couple of DSSU TNs that describes staging and cleanup:

http://www.symantec.com/docs/TECH66149 

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

View solution in original post

You also need to ensure one

You also need to ensure one backup images doesn't get close to 2TB.  If a client is larger than 2 TB it will abort with status 129. It will not "jump" to the next disk storage unit.

Instead of splitting the disk storage unit up in 2TB chunks it would make one big one. Less micro management more space if things get out of sync (like failed full backups). But - hey if it works for you that just fine wink

Highlighted

Yeah you're right the

Yeah you're right the frequency is set to 4 hours on every Disk Staging Schedule. But it looks like the disk will fill up to 90% and after that he'll begin with the job to the tape.

Therefore, the fact that staged images are not cleaned up when the DSSU is below high water mark does not represent a problem by itself. As soon as it grows above High Water Mark, a cleanup of staged images is triggered.

Doesn't this mean to lower the high water mark, will solve my problem, so the job from disk to tape will start earlier?

I've seen some documents saying that the high water mark default setting is 80% which sounds very good to me. 

As I've tried to explain

As I've tried to explain before:

1) Duplication to tape (staging) is triggered by the schedule in DSSU config.
Can you see that duplications are running as per the schedule?

2) Cleanup is triggered by HWM. (The default is 98 - you have changed yours to a much lower value.)
This will normally happen while backups are running.
Only images already duplicated successfully can be cleaned up.

Have you had a look at the 2 TNs in my previous post?
I believe DSSU principals and best practice is also described in NBU Admin Guide I.
(Links to manuals in Handy Links in my signature.)