Copy job to tape or straight from backup jobs?

[MCrown90]

Our backups are currently running like this: VMs>Backup>Backup Copy>Tape

I set it up like this because the tape library and the secondary backup storage are local to each other, reducing network traffic. I've noticed some weirdness, last week, the tape job copied two full backups of the copy job instead of one and it regularly had to wait 12h+ to start because the copy job hasn't finished merging.

Should I change it to backup from the primary backup storage is is there a better way to do what I'm currently doing?

Many thanks

[Shestakov]

Hello and welcome to the community!

... last week, the tape job copied two full backups of the copy job instead of one and it regularly had to wait 12h+ to start because the copy job hasn't finished merging.

Could you give more details please? Did it copy the same file twice or 2 different full backups?
Thanks!

[MCrown90]

Two different synthetic full backups.

[Shestakov]

Were both of the backups missing on tape or the job? What do you use as a source of the tape job, repository or backup copy job?
Thanks!

[MCrown90]

Nothing seems missing. The Source is the copy job.

[skrause]

I would not recommend using backups on repositories that are "remote" in relation to the tape server as your source. I had several issues with tape jobs craoking because the tape jobs can't handle network issues very well.