Confusion Around Backup Copy Jobs

[chjones]

Hi all,

I realize this may be a complete n00b question so I apologise ... please go easy on me

I am trying to use Object Copy jobs to keep the last 2 restore points of every VM we backup just incase we have a disaster. If there is a disaster we won't be going back several days, we just need the last day to get the business back working again. Depending on disk space I may increase this to a week or two, but for starters I want to start at 2 restore points.

I have deployed a couple of WAN Accelerators each with 2TB of disk for cache and added them to Veeam. I have created a few Object Copy Jobs and have them configured as such:

Copy Every: 1 Day starting at 6:00am
Virtual Machines: Added the Backup Job
Target Repository: Other datacentre's CIFS/Windows Store
Restore Points to Keep: 2
Data Transfer: Through built-in WAN accelerators
Schedule: Any time (continuously)

Using the above settings I would have assumed that as soon as a backup job finishes, any new restore points would be copied to the other datacentre automatically by the Backup Copy Job, and only the last 2 restore points would be kept.

The Backup Jobs all start at 6pm every weekday and sometimes dont finish until 2-3am, hence why I set the Copy Job for 6:00am. I'm not sure really how this time plays into the copy job as the schedule is Continuous, this just doesnt seem logically correct to me. Why do I set a start time if the copy is continuous?

Anyway, using these settings I left the job to run for a few days to see what would happen and I get varying results.

The first day, I got the error "Incremental copy was not process during the copy interval" for 15 of the 17 VMs in the job. The job ran for 8 hours from 6am and the next backup didnt start until 6pm so I dont get how it got this interval?

The second day seemed to work, except two VMs reported "latest restore point is already copied". Again, I dont get how this happened as there was definitely another backup of these VMs that occured overnight.

The third day, the two VMs that didnt succeed the day before did complete. But they were the only two VMs even listed in the job.

The whole idea of scheduling these jobs is doing my head in. In the Backup Job you can select "Configure secondary destinations for this job" and the job has this selected and has the Backup Copy job defined. I don't quite get why you can't just select this and when the backup job completes it auto triggers the Backup Copy. The schedules within the Backup Copy is confusing me. Can the backup just auto trigger the Backup Copy and just run till it completes the restore points from that backup?

If I ran the job a couple of times during the day it would be nice to have the Backup Copy auto run after each run of the backup job.

I am probably making things too complicated for myself and there is a very easy way to do this.

Cheers,

Chris

[tsightler]

Just set the interval to start at 6PM, the same time as the backups, then any new backups that show up after this time will automatically be copied as soon as the backup finishes.

[chjones]

Thanks Tom. Does it matter if the backup doesn't finished till the next day, will it still copy all restore points?

We are limited by the number of concurrent jobs writing to the Backup Repository so some jobs sit waiting for resources for some time. I don't mind letting them wait, they all complete well within our Backup Window.

[tsightler]

Yeah it's no problem. Setting the interval on the Backup Copy screen is a little like setting the SLA, i.e. how often you want a Backup Copy to be created. For example you may run backups every 24 hours, but only want an offsite copy once a day so setting the Backup Copy interval to 1 day will only copy one restore point each day, the first one created after the backup interval expires.

In your case, you simply want your nightly backups to be copied as soon as they are done so you can simply set the interval start time to coincide with the start of your backup jobs. The Backup Copy job will start a new interval and begin looking for a restore point that's created after 6PM, as soon as it appears and is available (i.e. the backup job completes), the copy of that restore point will begin.

WIth your previous setting the job would start looking for points at 6AM, wouldn't actually find anything until after the first job from 6PM ended, and probably couldn't finish the copies by the time the next 6AM interval starts. I think all you need to do is change the interval and you'll be fine, but please let me know if not.

[chjones]

That completely makes sense ... maybe you should write the User Guide

Thanks for your help, I'll give that a go and see how the job responds. Thanks heaps!