Backup Copy Jobs, Network Traffic Rules and Upload Streams

[NerfHerder]

Hi All,

We have a bunch of backup copy jobs that are getting backup data up to a cloud repo hosted by our Veeam partner. We have limited bandwidth for our organization (~100 Mbps Internet service), so I have set a network traffic rule to throttle this traffic. I also notice there is an option on the same screen for "multiple upload streams". I have left this at the default of enabled, with 5 streams per job.

All things considered, the backup copy jobs are working, though it can take a long time for them to finish for larger backups. We do on occasion have the copy jobs fail with errors indicating that the job couldn't finish before the interval ended or the connection was forcibly closed. I generally take these as just how life is with moving large amounts of data over the Internet. However, the last couple of days some of these copy jobs have failed with a new message like "resource not ready: cloud task".

I am not sure what this means? Is this generally an indication of a local VBR issue or a network issue or a target repo issue? Certainly, if this continues I will open a ticket, but figured I would look for comment here first.

Also - since we are still pretty new to Veeam, are there any general guidelines or best practices on successfully running copy jobs to a remote cloud repo like this?

Thanks much!
~D

[Natalia Lupacheva]

Hi,

Regarding the error message "resource not ready: cloud task" - it seems like an issue which has to be solved on the Service Provider site. I believe it's better to open a Support case.

As for best practices - you can check bp.veeam.com site to find some information about working with Veeam products.

Thanks!

[NerfHerder]

Thanks for the reply Natalia,

While reviewing the info you shared, I came across the following, and I think it means that I messed up with my initial implementation of Veeam here:
Specifying Backup Copy Interval for Periodic Copy Mode

When you configure a backup copy job, make sure that its backup copy interval covers the whole chain of restore points in the backup repository from which you plan to copy backups. The length of the backup copy interval has an impact on the algorithm of restore point selection. Veeam Backup & Replication copies only restore points that match the following criterion:

Time of restore point creation >= current time – backup copy interval

That is, if you have a backup chain whose earliest restore point is 1 week old, you need to set the backup copy interval to 1 week. If you set the backup copy interval to a smaller time interval, for example, 1 day, all restore points that are older than 1 day will fall out of the search scope, and Veeam Backup & Replication will not transfer such restore points. For more information, see Restore Point Selection

.

I have already encountered scenarios where:
* I have a backup job ("job a") that performs its first full, and then continues each day with its incremental. I have the job set to retain 7 restore points.
* A couple of days later I enable the backup copy job, using "job a" as its source, and a remote cloud repo as the destination. I specify an interval of 1 day and specify that it should retain 14 restore points
* The copy jobs starts, but does not complete copying all backup files to the cloud repo before the first interval expires. It - in fact - takes a few days / intervals to complete.
* My goal is to retain a smaller number of current restore points locally, and a larger number of restore points in the cloud repo (for longer term storage, etc.)

In the example I outline, if the backup copy job is able to get all the backup files in the first interval (1 day), then I am good. But since it takes multiple days to complete.... I am left with a broken backup chain on the cloud repo? And essentially have bad backups in the cloud repo?

Thanks,
~D

[Natalia Lupacheva]

Hi,

Yes, you are right, in your case it would be better to increase backup copy interval or try to seed your VMs.
You can find more details about seeding in this thread;
there are also some useful details in this thread.

Thanks!