No reversed incrementals with Backup Copy Job

[AlexB]

Hi

I wanted to use backup copy jobs for Veeam backups beeing copied to DR site. During testing I found that, although the original Job is configured with reversed incrementals (.vrb), the resulting backup copy is changed to forward incrementals (.vib). I could not find any Job Option where this could be changed. Is this behaviour by design or did I miss something?

AlexB

[chrisdearden]

Hi Alex,
its by design - the copy job will always produce a fwd incremental. When you hit your retention period , the last incremental will be rolled into the full , so its still an "incremental forever" process

[foggy]

Yep, backup copy job works with any backup mode as a source, synthetically building up a new restore point after reading incremental changes from the source repository. After initial synchronization, it is always incremental (full backup file (.vbk) is merged with the nearest incremental file (.vib) when retention applies).

[AlexB]

Thanks for clearing this up. Is there a reason for that? Wouldn't it be possible to have reversed incrementals, too?.

Another question: Is there a difference regarding reliability of a backup chain consisting of forward or reverse increments?

[dellock6]

Alex, since it's a forever incremental with synthetic creation of the VBK, there is almst no difference with a reversed incremental. They act the same way at the end, only forward incremental creates probably less I/O on the storage used for backup copy files.

About reliability, it's a forever incremental with synthetic creation of the full backup, just like reversed incremental. So I would say the reliability is the same.

Luca.

[Gostev]

Thanks for clearing this up. Is there a reason for that?

It is naturally a better way of storing data when dealing with WAN, which are typically unreliable and slow. This way, failed or incomplete incremental run will not leave the latest restore point in the inconsistent state, as it would be the case with reversed incremental way of storing the restore points.

Plus, there are additional technical benefits of storing the data this way as it comes to WAN acceleration specifically.