Two Backup Copy Jobs: Local and Cloud

[nick3young]

Hi All,

We currently have a 'main backup' and a 'backup copy job'. We are exploring the possibility of also backing our data up to the cloud using VEEAM Cloud Connect.

What are the best practices of running two backup copy jobs at once? I've still never managed to wrap my head around running one!! (with the regular "Backup Failed" emails I get when a backup copy job runs outside of it's window... which is apparently expected behaviour!?).

Since backup copy jobs are designed to run once the main backup is complete... how would it work with two?

Thanks
Nick

[foggy]

Hi Nick, you can have two backup copy jobs having the same backup chain as a source. There shouldn't be file access issues, since backup copy opens files for read, so provided they both fit into the backup window, you're good to go.

Regarding the issue you're seeing, did the initial backup copy job cycle also fail and the job still tries to catch up or is it not capable to complete even copying of incremental changes within the allowed window?

[nick3young]

Thanks for the response foggy. You helped me out with another post relating to the 'backup copy job'. Basically, the problem I have is that I occasionally get an error for the job saying that 'some VMs were not processed during the copy interval'. But then very shortly after... it completes successfully. Is there a way to remove this?

As for the two backup copy jobs - do I simply setup an entirely new backup copy job (exactly how I did for the first one) and just point it to a different target? Will this not hit performance then?

[Iain_Green]

Hi we run multiple copy jobs of backups without any issue.

Basically setup your copy job to point at the backup and then clone it and edit any settings you need to change. We use it so we have a local backup on site then one copy replicacting off site and then a a second copy job pointing at the same backup as the first which create GFS points and saves them at a third site

[foggy]

Basically, the problem I have is that I occasionally get an error for the job saying that 'some VMs were not processed during the copy interval'. But then very shortly after... it completes successfully. Is there a way to remove this?

Do you mean it completes during the next copy interval?

As for the two backup copy jobs - do I simply setup an entirely new backup copy job (exactly how I did for the first one) and just point it to a different target? Will this not hit performance then?

As an alternative, you can also use backups produced by your first backup copy job as a source for the second one. For this, you need to point the second one to a backup repository used by the first job to store its backups.

[nick3young]

Do you mean it completes during the next copy interval?

My main backup starts at '22:00' and typically finishes at around '23:30' (for my incrementals Mon to Thurs). I then perform an 'active full' every Friday which takes around 16hrs (finishes at around 15:00 on Saturday).
My copy interval is set as 'every 1 day' starting at '22:30' each night 7 days a week. On the Saturday, the backup copy job reports as running successfully for a full 24hrs (with all VMs processed), but then I get a fail every Monday at the start of the new interval (job starting at 22:30 on the Sunday)... but weirdly the 'fail' says the job has been running for only 22 minutes. It then completes successfully at around 2am *head explodes!*. I usually don't mind... Monday morning when I check the logs, I see the 'fail' but then a 'completed successfully' straight after. But it would be nice not to see that fail!

As an alternative, you can also use backups produced by your first backup copy job as a source for the second one. For this, you need to point the second one to a backup repository used by the first job to store its backups.

I like the sound of this. So I simply create a 'backup copy job' on the repository I use for my first backup copy? Would this be better than creating two backup copy jobs on my 1st backup server?

[foggy]

On the Saturday, the backup copy job reports as running successfully for a full 24hrs (with all VMs processed), but then I get a fail every Monday at the start of the new interval (job starting at 22:30 on the Sunday)... but weirdly the 'fail' says the job has been running for only 22 minutes.

You don't have anything to copy during weekends, since the source backup job doesn't run. Reminds me of this issue.

I like the sound of this. So I simply create a 'backup copy job' on the repository I use for my first backup copy? Would this be better than creating two backup copy jobs on my 1st backup server?

Yes. This would put some load off your primary repository.

[nick3young]

Here is the log for our backup copy jobs:

You don't have anything to copy during weekends, since the source backup job doesn't run

But I'm not sure that is strictly true is it? Since our full backup doesn't complete until 14:00 on a Saturday... this backup then needs copying. And since each copy interval finishes at 22:30 each night... this only gives roughly 10 hours to complete the 'copy job'. The full backup takes 16 hours... so the copy job should take at least that. If anything, I would expect to see an error on Saturday night to say 'jobs haven't completed within the interval'. But instead, as you can see from the image above, I get a 'success' every Saturday night (25/02 and 19/02). It also says that the copy job has been running for a full 24 hours... this confuses me. And then I get an 'error' for the Sunday (26/02) before a 'success' in the early hours of Monday morning (27/02).

[foggy]

And since each copy interval finishes at 22:30 each night... this only gives roughly 10 hours to complete the 'copy job'. The full backup takes 16 hours... so the copy job should take at least that.

The copy job copies changes only, so is not affected by the full backup occurred on source. Anyway, I suggest contacting support for a closer look at what happens.