Forum Discussion

Teslah's avatar
Teslah
Level 3
9 years ago

Backup exec does not delete old jobs

Hi, I have this problem from time to time in about every version of backup exec I can think of over the years. From time to time, backup exec sends an alert saying to free space on the disk where it is backing up. So I have to manually delete the old .bkf files, even though the job is configured like this;

 

Backup to a NAS, in the last case I had this problem, this NAS receives 1 backup per week, full, 2 servers backed up, 3 jobs in total (sql, exchange, data)

Job is configured to keep for 2 days

I would delete all the content of the directory where the backup is made, leaving only the 2 .cfg files. Then I would have a full wich will work, then next week when the second full would start, after a while, the job would still run until there is no free space (as I remember). So I manually delete the files of the first backup, go in alerts, respond to it, and the backup would countinue.

This kind of problem happens from time to time, in about every type of configuration (we have multiple clients), to clients randomly. It seems that most of the time, it would delete old backups by itself, sometimes it does not, I have to do it manually.

the only thing I could think of is when a problem appears during the backup and it does not complete, those .bkf files will not be deleted and over time, with other errors, these grow and take more and more space and are not deleted by backup exec automatically.

We always use removable storage. Maybe there is a setting I did not see somewhere or something I must do to make sure that backup exec will delete old job as needed to free space.

Any guidance would be appreciated.

 

Thanks

 

  • The title of your post is slightly misleading as your requirement is to automatic deletion of backup sets and not jobs.

    Have a look @ this post as well as to how DLM removes the backup sets - https://www-secure.symantec.com/connect/articles/automated-disk-management-and-data-retention-backup-exec-dlm-or-data-lifecycle-management

  • Do not delete the .bkf files from the disk storage.  This will leave orphaned backup set entries in the BE console.  When you need to delete backup sets, delete them from the console and BE will delete the corresponding .bkf files from the disk.

    You can use the registry setting at the bottom of this document to delete the last backup chain.

    http://www.veritas.com/docs/000085757

    Do note that if your backup job fails, you would be left with no backup set to do your restore.  It would be better to get a bigger storage than to use this registry setting.

  • Disk storage is managed by DLM. My article below may help you figure out what is wrong with your setup. https://www-secure.symantec.com/connect/articles/when-backup-sets-are-deleted-under-dlm Also, you need to leave you disks online for more than 4 hours so that they are present when the DLM grooming cycle runs
  • The title of your post is slightly misleading as your requirement is to automatic deletion of backup sets and not jobs.

    Have a look @ this post as well as to how DLM removes the backup sets - https://www-secure.symantec.com/connect/articles/automated-disk-management-and-data-retention-backup-exec-dlm-or-data-lifecycle-management

  • Hi, thanks both for your comment. In my last particular case, I think I understand what is happening. My disk does not have enough space for 2 full backups (about one and a half). Following VJware's article from the link, what I understand is that backup exec will never delete your last backup even if the media set is expired. I see that I have the option to tell backup exec to delete everyting that is expired, but since I have multiple different backup schemes in there, I don't want to do that.

    So the first backup goes fine, then the second one will continue until no more space left. I guess I understand that for good disk space management, you should always have a backup disk that can take at least 2 full backups.

    Maybe I did not understand everything correctly, but as of now, I'm still wondering how I can tell backup exec to delete the last full backup before doing it's new full backup automatically.

    The only idea I have now is to create and schedule a batch file that deletes all those .bkf files just before the new backup starts. This backup is just a secondary one we do once a week on a NAS, it's not that important but we still want it. You never have too much backup...:)

    If any other comments, feel free.

    Thanks all again for your input, I've learned a lot regarding media and disk management.

    Regards

  • Do not delete the .bkf files from the disk storage.  This will leave orphaned backup set entries in the BE console.  When you need to delete backup sets, delete them from the console and BE will delete the corresponding .bkf files from the disk.

    You can use the registry setting at the bottom of this document to delete the last backup chain.

    http://www.veritas.com/docs/000085757

    Do note that if your backup job fails, you would be left with no backup set to do your restore.  It would be better to get a bigger storage than to use this registry setting.

    • chathuraa's avatar
      chathuraa
      Level 3

      Hi Sir,

      I have this same problem in my backup exec environment. After the 2-week retention period backup(bkf files) keeps in my local computer backup partition. We only use tape backup. Please help me to resolve this.

      Please advise me on how to delete backup manually in exec console.

  • Hi, I how do I delete backup sets? The only options I have in the backup sets tab is;

    expire, retain, catalog, duplicate, verify, expiration, show dependent backup sets, restore

    Thanks

  • This one may also help understanding as although it is about RDX it discusses why we might need enough space for two complete backups

    https://www-secure.symantec.com/connect/blogs/backup-set-retention-and-expiry-handling-removable-rdx-etc-media