Backup job vs backup copy job (backup per machine files)

[BostjanUNIJA]

Hi.
I would like to understand what will happen in the scenario below.

Backup job in primary repository (backup per machine files currently NOT ENABLED).
Backup copy job in secondary repository (backup per machine files ENABLED).

Questions:

1. If we enable backup per machine files on primary repository, I guess the next scheduled run of the job will automatically run full backup, true? Because it needs to split backups per VM's?

2. How does the retention deal when you switch from per machines files not enabled to enabled? Does it still apply to a previous backup chain?

3. What happens to backup copy job sync (which already has per machine files enabled) after primary repository per machine files are enabled and active full backup is created, will it try to do a FULL BACKUP COPY JOB sync? Or will it be "smart" and copy only "some incrementals" since now primary and secondary repositories both will have per machine files ENABLED?

With best regards
B

[Dima P.]

Hello Bostjan,

Once you upgrade your primary job to a new backup format backup copy needs to be upgraded too.

1. If we enable backup per machine files on primary repository, I guess the next scheduled run of the job will automatically run full backup, true? Because it needs to split backups per VM's?

Correct, existing backup will be detached and new backup chain will be created, because there is no way to transform excising single full backup to a multiple separated per-vm backups.

2. How does the retention deal when you switch from per machines files not enabled to enabled? Does it still apply to a previous backup chain?

If daily retention was used then your backups will be a subject to background retention i.e. all 'orphaned' backups with a daily retention are removed from the repository when the latest restore point is retired.

3. What happens to backup copy job sync (which already has per machine files enabled) after primary repository per machine files are enabled and active full backup is created, will it try to do a FULL BACKUP COPY JOB sync?

You will need to upgrade backup copy chain too because it's not possible to have 'mixed' backup chains with old a new format. Keep in mind that backup copy job should be running on a new engine: mirror mode or periodic mode without intervals. Upgrade itself for the old per-vm chains is easy - you just need to upgrade the backup chain and that's it.

Or will it be "smart" and copy only "some incrementals" since now primary and secondary repositories both will have per machine files ENABLED?

If the existing backup copy chain was upgraded successfully the next job run will be incremental.

Take a look at this article for more details. Thank you!

[BostjanUNIJA]

Hi. Thank you very much for your reply.
When changing primary backup repository to: "per machine backup files" we receive a warning: "existing backup chains will not change their type until you detach the backups".
Is this something that we need to manually do in the job or will changes be reflected automatically after changing this on backup repository level?

Current backup copy jobs (which are in immediate mode) already have metadata for each VM backup, so I believe here no upgrade of chain will be needed...

With best regards

[Dima P.]

Bostjan,

Correct, you need to detach old backup chain manually.