Backup copy intervals for large fulls with small increments

[moodjbow]

Hi,

I know there are many topic discussing this, but still some small but important details are missing, for me to take a configuration decision.
We have a file server with 16+TB primary storage backups scheduled daily. On the weekend it takes full copies 38-48 hrs to complete. During weekdays the incrementals last only 15-20 mins.
We have the copy backup job against a slower secondary backup repository. The initial copy broke multiple times with the "Copy interval has expired" as described here: vmware-vsphere-f24/copy-interval-has-ex ... 33862.html and it tok it over 100 hrs to complete.
No, we didn't seed since we used the Hyper-V manual that is missing such information. Why didn't you guys update it? The topic "Creating Seed for Backup Copy Job" is only under vSphere manual.

Anyways we managed that, but the question still hang in the air:
I want to change the "Copy every" interval back to 1 day so that it catches my incrementals and copies them offsite. But what will happen if it takes longer than 24 hours to copy the latest restore point, say because of large data changes on the file server? The option "Read the entire restore point from source backup" is disabled, but nevertheless I can imagine that synthesizing can take longer than 24 hrs. What happens then?
To put it more straightforward: is the option "run until finished" as described here post208615.html?hilit=copy%20interval#p208615 soon to appear?

And one more improvement step would be if you make clear to us IT admins that the interval on the initial copy job should cover the whole chain of backups in the past that we want to copy.
I also made a mistake with smaller interval and now only hope that we don't have to restore from secondary repository too early information!

Thanks!

[Shestakov]

Hi Boian and welcome to the community!

No, we didn't seed since we used the Hyper-V manual that is missing such information. Why didn't you guys update it? The topic "Creating Seed for Backup Copy Job" is only under vSphere manual.

There is an article about seeding in Hyper-V user guide as well.

But what will happen if it takes longer than 24 hours to copy the latest restore point, say because of large data changes on the file server?

The job will start copying next increments, meaning you need to choose a proper interval.

And one more improvement step would be if you make clear to us IT admins that the interval on the initial copy job should cover the whole chain of backups in the past that we want to copy.
I also made a mistake with smaller interval and now only hope that we don't have to restore from secondary repository too early information!

Thanks for that feedback!

[foggy]

If the backup copy job is not able to copy all the changes within the specified synchronization interval, it will resume data transfer during the next sync interval, copying all the blocks required to build the new restore point corresponding to the VM latest state. Previous restore point though will be left in incomplete state.

[moodjbow]

Hello and sorry about the late reply: somehow I didn't get notified about your inputs.
Foggy, what you say with the incomplete state is scary. This means that we most probably have to delete the data from the secondary storage and seed with interims/auxiliary copy job.
OK, but even when we fulfill this, how can I make sure that the below three conditions are met:
The copy job should still check daily for the incrementals and copy them as needed
The copy job shouldn't stop until finished when the copy process takes longer than 24 hrs.
The copy job shouldn't error out when the parent backup job takes longer than 24 hrs. to complete.
OR enhancing the question from the first post, is it somehow possible to configure a copy job in such a way that it immediately starts after a successful backup job completion and runs until finishes, encompassing daily incrementals that take 15 minutes to complete with weekly fulls that take 40 hours or more?

[foggy]

The copy job shouldn't stop until finished when the copy process takes longer than 24 hrs.
The copy job shouldn't error out when the parent backup job takes longer than 24 hrs. to complete.

Both are not possible to achieve with the current implementation.

[moodjbow]

Thanks for clearing this out, Alexander. I hope in the near future this gets changed. Now we have to revise our backup copy policy to weekly. Not optimal.