Forum Discussion

Shug's avatar
Shug
Level 2
10 years ago
Solved

backups failing with error unable to allocate new media for backup, storage unit has none available(96) although there is tapes available

Hi Veritas Forum!

i'm having some problems with backups failing.

Bit of background. I switched offices in April Last year to become the infrastructure support for around 90 people, a part of this is the backup infrastucture.

we are using Netbackup 6.5.5 on a Server 2008R2 box with 2 tape libraries the main library has 2 HP ultrium 3 drive the secondary library has 1 ultrium 3 drive.

 

Since the weekend the backups have been failing with the error 'unable to allocate new media for backup, storage unit has none available(96)' although there are blank tapes in the library. I've checked the image has expired and also changed hte date for the tape expiring, I ran bpexpdate on all tapes (unsure if this was the right thing to do, this is what the last engineer told me to do in his notes) so i'm unsure what to check next.

I've had very limited experience with backup technologies so please take it easy on me.

 

Thanks in Advance.

Hugh.

  • run:

    veritas\netbackup\bin\goodies\available_media

    Post the output

    Which pool is your policy using?

    bppllist <policyname> -U

    Post the output

     

    Also please be aware NetBackup 6.x has been EOSL since Oct 2012 so you really need to upgrade to 7.7

7 Replies

  • does it bpexpdate: -deassignempty  command that you run on the tapes...? if this is the command, it will release the tapes if they do not have any valid images.. should not be a problem.

    i would like to know how you have confimred about the avaliablity of the tapes...?

    could you confirm below details?

    what is the Drive Density configured for 2nd library with one ultrium 3 drive.( does it Hcart/hcart2/hcart3) and what about the tapes density configured for tapes in the same library.. ( same info please for other library with 2 Drives)

    what is the library that was trying to used for failed jobs with EC 96?

    Does the tapes have the robot number with correct Densisty? and not frozen and suspended...?

  • run:

    veritas\netbackup\bin\goodies\available_media

    Post the output

    Which pool is your policy using?

    bppllist <policyname> -U

    Post the output

     

    Also please be aware NetBackup 6.x has been EOSL since Oct 2012 so you really need to upgrade to 7.7

  • I think this is the problem : " ....and also changed the date for the tape expiring" Tape expiration is not the same as image expiration (bpexpdate ). Tape expiration is a hardware expiration when tapes are old, faulty, taken out of operation. Change tape expiration date back to 'never'. PS: Expiring tapes with bpexpdate means you are expiring valid backups that are still needed for restores.
  • Hi All,

     

    Thanks for the replies. I managed to speak to someone from a different office who had a look for me and it turns out that all the available tapes were frozen for some reason.

     

    However i do have a new problem of the tape drives going down.....

  • You clearly have hardware issues... Faulty tape drives will also result in tapes being frozen.

    Seems your entire backup infrastructure is as old as your NBU software?

  • Hi Marianne,

     

    yes your correct, the Tape library/drive are very old however up until this weekend have been pretty reliable.

    I did have a problem with one of the tape drives a few months ago, however out hardware vendor replaced it and it was all working again. I've 'upped' both drives in the master library and the backups are running ok..... for now.

    I'm assuming that when the backups are running it's supposed to use both the tape drives at the same time? it only seems to be using 1 at the moment and the backups are stacking up because of this.

     

  • There are many things that need to be checked when all drives are not in use.

    1st thing is the STU selected in the policy - STU (robot) with one or two drives?

    Is the STU for 2-drive robot configured with 2 drives in STU properties?

    Are all the drives (still) UP?

    Are there any orphaned device allocations? (nbrbutil -dump)

    Enough free media available for all drives?

    What is the reason for jobs being queued? (See job details).

    You can see that this is a separate issue - not directly related to status 96 in this discussion.
    Maybe a good idea to start a new discussion for drive not getting used?

    PS:
    The 'available_media' report that revarooo asked for would show FROZEN state of media.
    Please mark his post as Solution?