Backup failure

[Timothy_Folsom]

I hope someone can help me. I am running 8.6 with my NT4 server. I have a Dell Powervault 122t SDLT autoloader attached. I have 3 backup jobs that are pointed towards that autoloader. I have 3 media sets created. Each one has a 4 week overwrite protection and an infinite append period. When my backup job executes, it grabs the right tape from the media set and the backup completed successfully. Once the tape gets to between 220Gig and 250Gig, the backup dies. When you look at the Robotics Library Inventory report you see that there was a 3 entered under the header of full, which means that there was a special state encountered. I have turned off Dr. Watson through the registry, replaced and updated to the latest firmware SCCI cards, replaced media, (and feel like float-testing backup exec). Once a job fails, I can create another job finishing up where the previous one failed, and it picks up the correct tape out of the media set and works fine. Why are my jobs failing?? Why am I geting the 3 when my tape almost reaches the max?

[Ken_Putnam]

No alerts within BackupExec?

Any Event Log entries that may apply?

Have you partitioned the library, or are the jobs just pointed to the library as a whole?

220GB on a 160/320 tape is about 1.375:1 and I usually figure about 1.3:1 as my rule of thumb, so I suspect that the tape is filling up.

[Timothy_Folsom]

I have not seen any alerts except that I need to insert overwritable media in the next slot, which already has media in it. No event logs that give any leads.

I have no partitions because I want more control over which slots are used for what. I just consolidated backups that were on 3 other servers with backup exec 7.2 to this one backup server with Backup exec 8.6 and the SDLT loader. I am still experimenting with how many tapes I am going to need for each job.

[Timothy_Folsom]

About the tape filling up....I thought that that might be the case, but I am wondering why the job is not moving to the next tape in the media set. Do you have to partition in order for the job to move to another tape in the series?

[Ken_Putnam]

Shouldn't have to. As long as the job is pointed to the library as a whole, when a tape fills up, Backup Exec should grab the next available scratch/overwritable tape. The option (Tools\Options\Overwrite, I think) "Use overwritabe media in target media set before scratch" will determine the order of use if more than one overwritable tape is found.

Just for grins. Before the job runs again, inventory all the tapes in the library. If they appear in blue, they are available for overwritem, if in black they are Write Protected.

Also, in ase you didn't already know, every time that you swap tapes in the library, you must inventory whatever slot has changed.