Forum Discussion

stefan_varga's avatar
15 years ago
Solved

multiple copies in netbackup

Hi,

we are using multiple copies defined in policy(inline copies) - backup to 2 separe libraries.

One of our STK library went down for maintanace.  The problem is, that new jobs are qeued but are waiting for resources.

11.1.2011 13:00:48 - awaiting resource l3-kv-1. Waiting for resources.

Reason: Robotic library is down on server, Media server: nj,
          Robot Type(Number): ACS(1), Media ID: N/A, Drive Name: N/A,
          Volume Pool: oracle, Storage Unit: kv-hcart3-robot-acs-1, Drive Scan Host: N/A,
          Disk Pool: N/A, Disk Volume: N/A

 

Is this normal behaviour? My expectation was, that multiple copies configuration paramater brings some sort of availability. If one library fails, the backup will start and continue with the second copy.

I would like to hear your opinion.

Thank you.

  • Clarification on setting [Continue if fail] for Inline tape copy (Multiple Copies)
    http://www.symantec.com/business/support/index?page=content&id=TECH61098

    The "continue if fail" only comes into play once resources have been allocated.

    "...

    Conclusion:
    The default behavior for Inline tape copy is this.  In order for the backup job to start, all resources must be made available.
    The [continue if fail] does not kick in until data starts flowing.

    Thus, if one copy fails to obtain resources (storage units, drive or tape), then the whole job will fail. Other copies won't start writing even resources are available for them.

    ..."

  • How was the 'Multiple Copies' configured? In the column "For each image if this copy fails" - what is selected for each of the images? 'Continue' or 'fail'?

  • Clarification on setting [Continue if fail] for Inline tape copy (Multiple Copies)
    http://www.symantec.com/business/support/index?page=content&id=TECH61098

    The "continue if fail" only comes into play once resources have been allocated.

    "...

    Conclusion:
    The default behavior for Inline tape copy is this.  In order for the backup job to start, all resources must be made available.
    The [continue if fail] does not kick in until data starts flowing.

    Thus, if one copy fails to obtain resources (storage units, drive or tape), then the whole job will fail. Other copies won't start writing even resources are available for them.

    ..."

  • Thanks for the clarification, Andy! I was actually wondering if it applies to jobs 'once started' but could not find the answer in the manual...

  • Are there any configuration parameter to avoid this behaviour?

    I would like to see the job started after some timeout, because in other case we need to create new policies or reconfigure existing one to handle library downtime.

    This would be very helpfull, because as I said I expected some sort of availability here.

    Thank you.

  • set the STU to "any available" - but that could mean that both copies run off the same robot.

  • I see Andy, but that is not an expected option.

    Thank you for your clarification.