Forum Discussion

Daniel_Be_'s avatar
Daniel_Be_
Level 4
10 years ago
Solved

NBU selection list wrong order

Hello all,

 

We've got the following situation:

We are doing a ndmp backup to tape with NetBackup 7.6.1.

 

The selection list looks like this:

NEW_Stream
Path1
NET_STREAM
Path2
NEW_STREAM
Path3
…
NEW_STREAM
Path30

 

What happens is that Path2 is backuped at the end (after Path30).

We had an issue with a Tape Drive - because of this Path2 failed first of all and at the end there was a second try of Path2 what succeded because another Drive was used.

Meanwhile the defect Tape Drive was replaced but unfortunately the behaviour that Path2 is backed up at the end is still there.

We have already cleaned the selection list, restarted NBU Services, typed in the selection list correctly (as it has been before).

Then the backup ran fine with the correct order - until the second backup.

After the second try everytime Path2 is backed up again after Path30.

We have already checked the "includes" file (\INSTALL\Veritas\NetBackup\db\class\Policyname) 

There the order is also correct.

 

Does anyone have an idea why NetBackup is doing this?

 

Thank you very much!

 

Best regards

7.6
Backing Up
Backup and Recovery
NetBackup
Windows Server (2003-2008)
  • Marianne's avatar
    Marianne
    10 years ago
    Curious to know why it is important that the paths be backed up in a specific order? One reason why timing could be pushed out is if you use Frequency based scheduling. If the schedule is for once every day, then it means every 24 hours. Subsequent backups will start 24 hours after the start of the previous successful backup. We call this 'schedule creeping '. To overcome this, make the schedule frequency smaller. Make it 1 hour more than the backup window. If the backup window is open for 12 hours, make the frequency 13 hours. The next day's schedule should be kicked off on time because enough hours have elapsed since the last successful backup. I have posted David Chapa's article about frequency scheduling over here: https://www-secure.symantec.com/connect/articles/netbackup-frequency-based-scheduling-1 It explains how manual backups and/or 'schedule creeping' will influence normal schedules and how to overcome/prevent it.

4 Replies

  • GulzarShaikhAUS's avatar
    GulzarShaikhAUS
    Level 6
    10 years ago

    If I understood you correctly, this started happening after the path2 failed sometime. If so you can actually recreate the whole policy and try. 

    Also if you have multiple tape drives you can make use of "allow multiple data stream" options and can run simultaneous backups. This can also change the behavior.

  • Marianne's avatar
    Marianne
    Level 6
    10 years ago
    Curious to know why it is important that the paths be backed up in a specific order? One reason why timing could be pushed out is if you use Frequency based scheduling. If the schedule is for once every day, then it means every 24 hours. Subsequent backups will start 24 hours after the start of the previous successful backup. We call this 'schedule creeping '. To overcome this, make the schedule frequency smaller. Make it 1 hour more than the backup window. If the backup window is open for 12 hours, make the frequency 13 hours. The next day's schedule should be kicked off on time because enough hours have elapsed since the last successful backup. I have posted David Chapa's article about frequency scheduling over here: https://www-secure.symantec.com/connect/articles/netbackup-frequency-based-scheduling-1 It explains how manual backups and/or 'schedule creeping' will influence normal schedules and how to overcome/prevent it.
  • Daniel_Be_'s avatar
    Daniel_Be_
    Level 4
    10 years ago

    Hello Marianne,

    thank you very much for your answer. 

    We've create a completely new policy based on the desired backup sequence. This new policy works fine. 

    Now the question is: Why does NBU the faulty Volume at the following starts of Policy always at the end,although the order of the resources is still the same in the policy.

    And the other question is: Do we have to create a new policy every time after we had an issue with a tapedrive? Or is that a bug in Netbackup 7.6.1?