Feature Request: Remove interval requirement for Backup Copy

[souperstar]

I've tried, a lot, to get Backup Copy Jobs to sync my direct attached storage repository to removable disks that go offsite daily. All I really need is: as soon as a backup job finishes, start copying to the secondary destination defined in the Backup Copy Job. The current "interval" scheduling and processing is just not helpful to achieve Backup Copies to rotated media. If the scheduling/"interval" requirement were removed, I think that would solve a lot of issues.

For example, say I have a 2TB VM backing up at 6 PM. It takes a while. Say it's done at 8:00 PM. Now, there is no way to be "sure" when this job will finish. It may finish early some days, it will certainly finish later on others. So, when am I supposed to schedule the Backup Copy Job interval? I'll just schedule it at 7:00, seems safe. Now, this all hinges on some arbitrary interval. What if my first Backup Job starting at 5 is running late so my 6:00 job is pushed back to 8 PM. My interval doesn't sync my backup! At least, that's what I experienced and support confirmed.

I'm still not sure what secondary destinations are for, if not exactly for what I want: make a copy of this backup from repo A to repo B when the job is done. Why do I need an "interval"? I've asked, I still don't have an answer.

Plus, if I set the interval to every four hours for a Backup Job with multiple VMs, and the VMs don't get sync'd within that four hour window, then I'm hurting bad. The job doesn't finish on its next interval, it just sits there saying it's waiting for new restore points for the VMs it missed. How is that a copy? It didn't copy! It a) stopped before it finished and b) didn't resume.

The whole thing seems illogical, but maybe I'm using it for an unintended purpose. I was informed by support that there are no such interval restrictions on Tape Jobs.

So please, please remove the interval requirement on Backup Copy Jobs. The Backup Copy Jobs are a great idea, I just need them to get a copy of a backup on rotated media immediately following a successful Backup Job, and work at doing so until it is finished. That's it!

If you would like to read more about my environment and my setup, it is here:
http://forums.veeam.com/veeam-backup-re ... ml#p154953

If you would like to see what support and I have tried in order to succeed, here is the case # 00985902 (closed).

That is all, thank you for the great product Veeam!

[JaxIsland7575]

Chris,
Just to throw in my 2 cents, I use the Backup Copy Interval like this: Backup starts 1800, Backup Copy Starts 1805. I set it to run every day. This way the new interval beings 5 minutes after the backup job starts. So the backup job has 23 hours 55 minutes to complete before being missed, this setup has always worked for me. So say my 1800 Backup job completes at 1930 the Backup Copy recognizes the new backup file and starts approximately 2 minutes later and does the copy.

Cheers!

Edit: On another site that was smaller I had 6 Backup jobs starting at 1900 each night. They start at the top of each hour so 1900 to 0000. I set the a single Backup Copy job to look at all 6 backup jobs and set that for each day starting at 1905. This has always captured the new files after the backups and copied them to the secondary location.

[souperstar]

But when your 1700 backup job runs late and your 1800 doesn't start until 1810, then what? You'll be right where I was, without the Backup Copy.

EDIT: Which is why they suggested a shorter interval, which would work for smaller data but it takes hours for my copies to complete. The shorter interval would thus not finish syncing all VM's in a single interval, but then wouldn't finish when the interval started over and the exact same data was still present. I could move my Copy Job start time further out, but then I run the risk of them not being finished by 0700ish when they rotate them.

Thank you for your suggestion, I'm still very interested in using Backup Copies.

[JaxIsland7575]

But when your 1700 backup job runs late and your 1800 doesn't start until 1810, then what? You'll be right where I was, without the Backup Copy.

If I have a Backup Copy job set to start at 1802 for the 1800 Backup job, it still has 23 hours and 55 minutes to finish and start the copy job. So even if the 1800 job gets pushed back to 1900 I still have almost 23 hours for it to complete.

[dellock6]

And it's mode is "continuous" so it will catch the new restore point as soon as it's into the primary backup.

[souperstar]

And it's mode is "continuous" so it will catch the new restore point as soon as it's into the primary backup.

I agree, and this makes sense, however it wasn't my observation... I have the case posted and multiple technicians webex'd and verified. Veeam was somehow "stuck" or "broken" and we had to kick everything off manually in order to get it unstuck. But that didn't resolve everything, because then Veeam didn't overwrite the data that was on my removable disks anymore and started over on them. Filling them up and again, not giving me a successful copy.

Too unreliable for me. At the rate you guys are incrementing versions, I'll wait for 9 and hope this gets put in. In the meantime I have closed the case and went back to robocopy.

[Gostev]

It's too late to put in v9 because it is a major conceptual change, but we will consider this for the future versions. However, may I ask that in future you don't agree to close the case until the issue is fully resolved for you. Otherwise, it doesn't get escalated into R&D and thus remains in the product...