Large VMs and Backup copy

[justin23]

I'm having issues getting backup copy jobs to completely copy full backups of large VM's (8TB Exchange and a 13TB file server).

The backup copy jobs run daily at 8am. I have a copy window of 2am until 8pm (Sunday to Friday). I find the performance suffers if I run Backup copies when the backups are trying to run.
We are keeping a lot of backups like 8 weeklies, 24 monthlies and 7 yearlies.

The issue si the larger servers the data doesn't seem to be copying across in time. I thought it was just pickup from where it left off ate the next window, but appear not. Instead all I'mm getting in the report is "copy Interval has expired Incremental copy was not processed during the copy interval"

How should my backup copy jobs be configured? All I want is for the weekly full backups to be copied and tagged correctly for retention. I don't care if it take several days to do so, but I'm currently not retaining customer data. I've logged a job regarding this in the past and just got given links to how Backup Copy jobs works. I've since redesigned all the backup jobs and backup copy jobs but I'm still finding the same issue with backup copy jobs. It can't be this hard!

How should I configure a backup copy job to ensure I get my weekly backups copied and marked for retention correctly. I don't care about incremental backups during the week.

Thanks for any assistance people can provide.

[Mike Resseler]

Hi Justin,

If I understand correctly, you have a bad performance when backup is running while the backup copy job is busy. Is that correct? BCJ will not work when the restore point is locked by the backup job. (For example when the backup job is doing a transform on the chain)


Can you also check from the last BCJ what your bottleneck is? That might give us an idea what is going on

Thanks
Mike

[Shestakov]

Hi Justin,
How often do the source jobs run(daily/weekly)?
As Mike said, source job locks backup files so backup copy cannot copy them. As a workaround you may separate primary and copy jobs' backup windows.
Thanks!

[justin23]

The performance isn't really the issue here. The issue is I can't get a successful backup copy job to complete of full backup large VMs. I'm even just copying from one store to another on the same device to try and complete it and its not completing.
Even if I opened the schedule up to the full 24 hours, the large VM doesn't complete in time and I'm getting inconsistent copies.
On small servers I've got 13 restore points, yet on these two big servers i'm lucky to have one and i'm not even sure its valid. The logs just say the interval ended before the backup copy job completed. I'm really after a best practice or best way to configure Backup Copy jobs to ensure I get the full backup job copied and retained as a weekly, Monthly or Yearly. I thought the job would just continue, but apparently not.

[Mike Resseler]

Justin,

What is your backup copy interval? (https://helpcenter.veeam.com/docs/backu ... tml?ver=95)
Most probably (at least I am guessing here) the first backup copy from those large VM's does not make it (we have seeding for that)

[justin23]

So I opened the backup copy jobs to the full 24 hours and the interval is set to run every day at 8am. Still have the same issue. I was thinking of running it once a week with a 7 day interval, but if the full backup isn't complete, it will just pick up the previous incremental, or if another incremental has occurred it would just pickup that incremental, so its a bit too risky.

I only care about copying the full backup every week, but it doesn't seem to be an option in backup copy jobs and it would be a great option (and one that should be pretty easy to implement). I could then run 7 day intervals and just do the copies during the week.

So I'm thinking the seeding could be part of my issue. When I look at the files I'm seeing a 33.5 MB monthly backup, for a server that is 13 TB. I'm also working on splitting that server into two smaller servers, but would still like to understand the root cause of my issue.

[justin23]

regarding seeding. I've found this post
vmware-vsphere-f24/re-seeding-on-copy-j ... ml#p110449
However as the two datastores are in the same datacentre I don't see the point of an intermediate point. Could I just Run an Active Full on the backup copy job and set the windo to say a few days so it will complete. will that seed the backup copy correctly?

[Mike Resseler]

If I understand it correctly, if the BCJ makes it in that timeframe then it will indeed

[justin23]

No worries, I've set a 7 day window and I'm running active full backup on the BackUp Copy Jobs that have the big VM's. Is it better to have seperate Backup copy jobs for large VM's or just have them in one job with all the other VMs?

[Shestakov]

Give it a try, if the job works fine with all VMs, all is good.
If the job will take too long and bottleneck can't be expanded, split VMs to 2 jobs.